From Vocabulary to Voice: How Generative AI Shapes Arabic Learning for Non-Native Speakers

From Vocabulary to Voice: How Generative AI Shapes Arabic Learning for Non-Native Speakers | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article From Vocabulary to Voice: How Generative AI Shapes Arabic Learning for Non-Native Speakers Ahmed Khouli, Abed Alkhaleq Esa, Bilal Hamamra, Suhad Nierat, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9249983/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract This study examines how generative AI tools reshape Arabic learning for non-native adult speakers, focusing on a diverse cohort of Greek-speaking learners distributed across Saudi Arabia, the UAE, Greece, and other European countries. Through an open-ended questionnaire and thematic analysis, we explored learners’ experiences, adaptive support features, and perceived challenges associated with AI-driven applications such as ChatGPT, Duolingo, and Google Translate. Findings reveal that AI significantly enhances motivation by simulating culturally rich, low-anxiety conversational scenarios and delivers real-time personalization in vocabulary reinforcement, pronunciation practice, and grammar scaffolding. However, persistent limitations include translation inaccuracies, inadequate handling of regional dialects, and risk of cognitive overreliance when learners depend on AI for routine corrections. The most effective learning environments combined generative AI with instructor-mediated discussions, leveraging AI’s strengths while preserving human oversight for nuanced cultural and linguistic guidance. Implications underscore the need for dialect-aware AI models, equitable infrastructure investment, and continuous educator training. This research offers empirical insights for ethically integrating AI into hybrid and virtual Arabic programs to support personalized and culturally responsive language acquisition. Generative AI Arabic language learning Non-native speakers Adaptive feedback Cultural engagement Introduction In recent years, artificial intelligence (AI) has emerged as a transformative force in language education, offering unprecedented opportunities to personalize, adapt, and enhance the learning experience for students across diverse linguistic and cultural contexts. AI-powered tools such as intelligent tutoring systems (ITS), natural language processing (NLP), automated writing evaluation (AWE), and chatbots are increasingly used in second and foreign language instruction to support engagement, adaptivity, and linguistic proficiency (Son, Ružić, & Philpott, 2023). These technologies provide learners with real-time feedback, tailored content, and immersive environments that simulate authentic communicative situations, thereby addressing many of the limitations of traditional, one-size-fits-all instruction. The potential of AI in Arabic language education has gained particular attention due to the language’s inherent complexity, including its rich morphology, diglossia, and challenging phonetic features (AlAfnan, 2025). AI-driven tools are especially useful for supporting non-native speakers in distinguishing between similar Arabic sounds, improving intonation, and developing more accurate pronunciation models—tasks often difficult to achieve through conventional instruction alone. Moreover, adaptive AI systems can respond to individual learners’ progress by modifying content and pacing, thereby creating a dynamic and inclusive learning experience (Pokrivčáková, 2019). Despite this potential, the application of AI in teaching Arabic to non-native speakers remains underexplored compared to its use in English or other widely taught languages. While studies have emphasized the transformative impact of AI on Arabic language acquisition for native speakers (Alsaied, 2024), research focused on adult learners from diverse cultural and professional backgrounds, such as Greek-speaking students learning Arabic across multiple countries, is still limited. These learners, with motivations ranging from professional development to intercultural communication, often face additional challenges such as dialect variation, formal-informal language distinctions, and pronunciation accuracy—all areas where AI can provide meaningful support (Son et al., 2023; Pokrivčáková, 2019). Furthermore, scholars have called for a more integrated and theory-driven exploration of how AI can be implemented in foreign language education, not only to support technical aspects of learning but also to foster emotional engagement and cultural sensitivity (AlAfnan, 2025). This is especially relevant in Arabic instruction, where learners benefit from emotional connection and context-based use of the language. Against this backdrop, the present study investigates how AI tools contribute to enhancing engagement, adaptively, and proficiency in Arabic among non-native speakers. Focusing on a diverse cohort of Greek-speaking learners located in the Gulf region, Greece, and other European countries, this research aims to provide empirical insights into how AI technologies mediate the experience of learning Arabic in virtual, culturally diverse settings. Research problem While interest in Arabic language learning is growing, non-native adult learners—especially those from diverse cultural and professional backgrounds—continue to face challenges in mastering Arabic’s complex phonology, grammar, and script (AlAfnan, 2025; Alsaied, 2024). Traditional instruction, often delivered through fixed online curricula, lacks the personalization needed to address varied learner needs, particularly in multicultural contexts like that of Greek-speaking learners across Europe and the Gulf. Artificial intelligence (AI) tools—such as intelligent tutoring systems, catboats, and adaptive learning platforms—have shown promise in enhancing engagement, adaptively, and language proficiency in second language education (Son et al., 2023; Pokrivčáková, 2019). However, their application in Arabic learning remains limited and fragmented, often hindered by infrastructural gaps, insufficient teacher training, and ethical concerns (Pokrivčáková, 2019). This study addresses the gap by exploring how AI can be effectively used to support Arabic language learning for non-native speakers in diverse, hybrid learning environments. Research purpose The purpose of this study is to explore how artificial intelligence (AI) can enhance engagement, adaptively, and linguistic proficiency in Arabic language learning among non-native speakers in diverse cultural and geographic contexts. Focusing on a group of Greek-speaking adult learners enrolled in Arabic programs across Saudi Arabia, the UAE, Greece, and other European countries, the study examines how AI tools address individual learning needs, support pronunciation and phonetic accuracy, and personalize instruction across varying proficiency levels (A1–C2). Given the challenges associated with Modern Standard Arabic and the learners' wide-ranging motivations—professional, academic, cultural, and humanitarian—the study investigates the role of AI in simulating real-life communication, reducing language anxiety, and fostering emotional and cognitive engagement. Ultimately, this research seeks to generate empirical insights into the effective and ethical integration of AI in Arabic language education for non-native learners, particularly within hybrid and virtual learning environments. Research questions 1. How do non-native learners describe their experiences with learning Arabic using generative AI tools? 2. In what ways do generative AI tools provide personalized and adaptive support for Arabic language learners? 3. What challenges do learners face when using generative AI tools for Arabic language learning? Related studies Gamification offers non-native Arabic learners interactive, game-like experiences—using points, badges, leaderboards, and streaks—to lower anxiety and boost motivation (Almelhes, 2024 ; Kotob & Ibrahim, 2019 ; Sahrir & Alias, 2011). Smartphone-based games have been shown to enhance speaking skills (Kenali et al., 2019 ), and touchscreen apps improve letter-form recognition (Al Hejaili & Newbury, 2023). Culturally relevant game content further strengthens comprehension and retention (Saleh, Arifin, & Hanefarezan, 2022). Platforms like Duolingo deliver immediate rewards that reinforce foundational vocabulary and grammar, though they may struggle with advanced communicative tasks (Loewen et al., 2019 ). Adaptive AI systems mark a significant evolution from static gamified approaches by offering real-time personalization that caters to individual learning needs. These systems continuously analyze learners’ performance and automatically adjust instructional content, providing a tailored learning trajectory that is particularly vital for mastering Arabic’s unique challenges in pronunciation, grammar, and script memorization (Putri et al., 2021; Riwanda et al., 2021). Nasaruddin ( 2024 ) reported that AI-driven platforms, including those integrating ChatGPT, can generate customized exercises that evolve as learners progress. This adaptive feedback mechanism not only maintains high levels of engagement but also supports self-regulated learning, allowing learners to focus on their specific weaknesses—a critical improvement over one-size-fits-all gamification models. Intelligent tutoring systems (ITS) and automated writing evaluation tools form the backbone of many AI-driven approaches in language education. ITS platforms simulate one-on-one tutoring by using machine learning algorithms to provide immediate, adaptive feedback on tasks such as sentence construction and vocabulary drills (Putri et al., 2021). Research shows that these systems help learners internalize complex grammatical structures by adjusting task difficulty in real time (Nasaruddin, 2024 ). Similarly, automated writing evaluation systems use natural language processing (NLP) algorithms to detect errors in syntax, morphology, and style in learners’ Arabic texts, offering instantaneous corrective feedback that enhances writing fluency and accuracy (Lu, 2019 ; Zhao & Zhang, 2019 ). Together, these tools reduce the workload on instructors while fostering a self-directed learning environment where learners can repeatedly practice and refine their language skills. Conversational agents and chatbots have emerged as a promising means of providing interactive dialogue practice. These AI-driven tools are designed to simulate natural conversation and offer a safe, low-pressure environment where learners can experiment with new vocabulary and sentence structures (Fryer & Carpenter, 2006 ; Thompson et al., 2018). Nasaruddin ( 2024 ) noted that such systems increase learners’ willingness to engage in spontaneous conversation, while studies by Dokukina and Gumanova ( 2020 ) confirm that chatbots help reduce communication anxiety by eliminating the fear of negative judgment. In the context of Arabic—a language with significant phonological and syntactic complexities—such immediate feedback is invaluable for improving fluency and pronunciation. Speech recognition features, now increasingly integrated into these systems, further allow learners to practice oral skills and receive real-time pronunciation corrections (Fathi et al., 2024 ). ChatGPT, a state-of-the-art conversational agent developed using deep learning and advanced NLP techniques, has drawn considerable attention for its potential in language education. ChatGPT is capable of generating contextually appropriate and coherent responses, making it ideal for generating tailored educational materials, interactive practice dialogues, and immediate corrective feedback (Nasaruddin, 2024 ). Its ability to simulate “human-like” conversation offers non-native speakers an “anxiety-free” space to practice and build confidence in their communicative skills (Fryer & Carpenter, 2006 ; Thompson et al., 2018). Although some studies have identified limitations—such as occasional generic responses or a lack of depth in sustained interactions—research suggests that these shortcomings can be mitigated by incorporating structured, teacher-guided activities alongside ChatGPT use (Kasneci et al., 2023 ). This blended approach leverages ChatGPT’s strengths while ensuring that cultural nuances and advanced linguistic subtleties are adequately addressed. The successful integration of AI technologies into Arabic language instruction depends heavily on the digital competencies and pedagogical readiness of educators. Studies indicate that teachers who are well-prepared and maintain a positive attitude toward technology are more effective in implementing AI-based tools (Al-Bulushi & Al‐Issa, 2017 ). Comprehensive professional development programs are critical; Pokrivcakova (2019) argues that such training should address both the operational aspects of AI systems and strategies for integrating automated feedback into the broader pedagogical framework. Research by Stošić and Guillén-Gámez (2024) and Nguyen ( 2024 ) underscores that teacher training not only improves educators’ technological literacy but also enhances student engagement and learning outcomes. Blended learning models that combine AI-driven activities with traditional, teacher-led instruction have been shown to yield the most positive results, as they allow educators to contextualize digital interactions with cultural and situational insights. Despite the many benefits of AI-driven language instruction, significant challenges remain. Robust technological infrastructure is essential, yet many educational institutions—particularly in under-resourced settings—struggle to afford the advanced hardware and high-speed internet necessary to support these systems (Azmi & Zakaria, 2020 ; Wulantari et al., 2023 ). Moreover, AI systems often face inherent limitations in capturing the full nuances of human language. While automated tools effectively correct grammatical errors, they sometimes fall short of providing the contextual, cultural, and stylistic feedback required for proficient Arabic communication (Zhao & Zhang, 2019 ; Fryer & Carpenter, 2006 ). Another critical challenge is the sensitivity of AI systems to input accuracy. Minor spelling or grammatical errors can trigger inappropriate feedback, potentially frustrating learners and detracting from the overall learning experience (Coniam, 2014 ). Ethical considerations—including data privacy, algorithmic bias, and equitable access—further complicate the integration of AI into language education. Pokrivcakova (2019) cautions that without rigorous ethical safeguards, these technologies may inadvertently reinforce existing inequities or foster culturally insensitive practices. Addressing these issues requires not only technological refinement but also the development of comprehensive policies and infrastructure investments. Comparative assessments of IT tools in language education offer valuable insights into the relative strengths and weaknesses of different approaches. Stošić and Guillén-Gámez (2024) conducted a comparative analysis of several IT tools—including Duolingo, Rosetta Stone, Edmodo, and Tandem—and found that each tool offers unique advantages that address different facets of language learning. For example, Duolingo’s gamified approach excels in boosting engagement and facilitating vocabulary acquisition for beginners (Loewen et al., 2019 ), whereas Rosetta Stone’s immersive framework is particularly effective in developing listening and pronunciation skills (Lord, 2015 ). Collaborative platforms like Edmodo foster group interaction and critical thinking, and Tandem, which connects learners with native speakers, has been shown to improve speaking fluency and intercultural competence. These findings suggest that no single tool can address all aspects of language learning; rather, a multimodal, hybrid approach is necessary to meet the diverse needs of non-native Arabic speakers (Dokukina & Gumanova, 2020 ; Woo & Choi, 2021 ). Emerging technologies such as augmented reality (AR) and virtual reality (VR) offer exciting new possibilities for Arabic language instruction by creating immersive learning environments that simulate authentic cultural contexts. Valencia et al. ( 2022 ) and Panagiotidis ( 2021 ) have shown that AR/VR applications can enhance both linguistic competence and intercultural understanding by providing realistic simulations of everyday interactions. Although research specific to Arabic is still in its early stages, promising findings from other language contexts (Wang et al., 2019; Viktoria et al., 2018) suggest that these immersive technologies can be effectively adapted to teach Arabic. Mobile learning applications further extend the accessibility and flexibility of language education. Gafni et al. ( 2017 ) and Talan et al. ( 2024 ) have documented that mobile platforms—by combining gamification with AI-driven adaptive learning—enable learners to access personalized content anytime and anywhere. These applications break down geographical and temporal barriers, allowing non-native Arabic speakers to engage with learning materials on the go. Moreover, mobile learning fosters continuous, self-directed study, a feature that is particularly beneficial in modern educational contexts where flexibility and personalization are key. A holistic, multimodal approach that integrates various IT tools and emerging technologies is essential for addressing the complex challenges of teaching Arabic to non-native speakers. Evidence from the literature suggests that gamified platforms are most effective for early-stage vocabulary acquisition and engagement, while immersive systems like Rosetta Stone bolster receptive skills through context-driven practice. Collaborative tools such as Edmodo enhance communicative competence by fostering group interaction, and real-world conversation platforms like Tandem significantly improve speaking fluency and intercultural competence (Stošić & Guillén-Gámez, 2024; Dokukina & Gumanova, 2020 ). Furthermore, the integration of AR and VR could provide the cultural context and immersive experiences necessary for comprehensive language acquisition, complementing the adaptive strengths of AI-driven systems (Panagiotidis, 2021 ; Valencia et al., 2022 ). To fully harness these benefits, continuous teacher professional development is imperative. Educators must receive ongoing training to stay current with technological advancements and to integrate these tools effectively into a blended instructional model (Al-Bulushi & Al‐Issa, 2017 ; Pokrivcakova, 2019). Finally, addressing infrastructural and ethical challenges—by investing in robust technology, developing rigorous privacy policies, and ensuring equitable access—is essential for the sustainable integration of AI in Arabic language education. Context of the study A comprehensive study of AI in Arabic education involves 19–59-year-old Greek learners spread across Saudi Arabia, the UAE, Greece, and other European countries. Their proficiency ranges from A1 to C2, creating a culturally and linguistically diverse learning environment. Instruction follows a fixed textbook aligned with the curriculum and is delivered via Zoom: the syllabus appears onscreen, discussions occur entirely in Arabic, and interactive exercises simulate real-world scenarios. Learners’ motivations vary: some seek career advancement—especially in Gulf job markets—while others pursue cultural understanding, academic interests in Middle Eastern studies or comparative religion, or personal reasons such as family ties. In Greece, many students are police, military, or human-rights workers aiming to communicate with migrants from Syria, Palestine, Iraq, and Libya; despite diverse dialects, they often focus on Modern Standard Arabic as a common linguistic medium. AI tools create virtual environments that bridge abstract learning and lived experience, fostering emotional engagement with language through realistic simulations. They also offer accurate pronunciation models and feedback on subtle phonetic contrasts—soft versus emphatic, voiced versus voiceless, plosive versus non-plosive—that non-native speakers frequently confuse. By combining these technological affordances with interactive, instructor-led activities, AI enhances both the cognitive and affective dimensions of Arabic acquisition. Methodology We adopted a qualitative research approach to investigate how generative AI tools are being used to support Arabic language learning among non-native speakers. A qualitative approach is particularly appropriate for exploring how individuals experience and interpret emerging educational technologies within their specific learning contexts (Mohajan, 2018). To guide our inquiry, we employed a case study design, which allows for an in-depth examination of a real-life phenomenon within its natural setting. This design enabled us to explore the integration of generative AI in Arabic language instruction from the perspective of diverse learners across cultural and geographical contexts (Yin, 2003). Participant The study involved 25 Greek-speaking adult learners of Arabic, ranging in age from 19 to 59 years. This diverse cohort was geographically dispersed across Saudi Arabia, the United Arab Emirates, Greece, and other European countries, reflecting a rich blend of cultural and professional backgrounds. Participants exhibited varying levels of Arabic language proficiency, from beginner (A1) to advanced (C2), creating an inclusive learning environment that accommodated a wide range of linguistic abilities. Instruction was delivered via Zoom using a standardized textbook aligned with the formal Arabic language curriculum for non-native speakers. Classes were conducted entirely in Arabic and emphasized interactive, real-world application through exercises and discussions. Participants’ motivations for learning Arabic were multifaceted. Some aimed to enhance career opportunities, particularly in Gulf countries, while others pursued the language for academic, cultural, or personal reasons—such as deepening their understanding of Arab societies, supporting human rights work, or connecting with Arab family members or partners. In Greece specifically, many learners were military personnel or police officers seeking to communicate more effectively with migrants from Arabic-speaking regions. Recruiting the participants Participants in this study were selected based on specific criteria aligned with the research objectives. All participants were non-native speakers of Arabic, with Greek as their first language, and were either currently enrolled in or had recently completed formal Arabic language courses designed for non-native speakers. They ranged in age from 19 to 59 years, ensuring a broad representation of adult learners. To capture diverse experiences with Arabic language education, participants were drawn from various geographic locations, including Saudi Arabia, the United Arab Emirates, Greece, and other European countries. Their proficiency in Arabic varied widely—from beginner (A1) to advanced (C2)—which allowed the study to examine how generative AI tools support learners at different levels. The selection also considered the participants’ motivations for studying Arabic, which ranged from professional development and academic interest to humanitarian work and personal or cultural reasons. Although prior use of generative AI tools such as ChatGPT, Google Translate, or Duolingo was not a strict requirement, participants with relevant experience were prioritized to ensure meaningful insights into the role of AI in Arabic language learning. Data collection To collect qualitative data for this study, we used an open-ended questionnaire designed to capture participants’ lived experiences with using generative AI tools in Arabic language learning. This method allowed for in-depth exploration of participants' perceptions, practices, and challenges, as they articulated their responses in their own words without constraints. Open-ended questions are widely recognized for their ability to elicit rich, detailed insights into individuals’ experiences and attitudes (Creswell, 2014; Harlacher, 2016). According to Thorndike and Thorndike-Christ (2010), open-ended items empower participants to express their thoughts freely, enabling researchers to gather nuanced information on complex phenomena. The questions used in the form were developed based on findings from prior research on technology use in education, particularly the works of Joo et al. (2016), Çoklar et al. (2017), Özgür (2020), and Christian et al. (2020), and were adapted to the context of generative AI and language learning. The final version of the form was written in Arabic and distributed to a diverse group of Greek-speaking adult learners of Arabic, all of whom were over the age of 18. The questionnaire consisted of three sections. The first section provided participants with an overview of the study and outlined their rights as voluntary participants. The second section collected demographic information. The third and main section contained ten open-ended questions focusing on the participants' experiences with generative AI tools—such as ChatGPT, Duolingo, and Google Translate—and how these tools have influenced their Arabic language learning journey. The questions addressed key areas such as AI-assisted vocabulary development, grammar support, pronunciation practice, user engagement, perceived challenges, and comparisons with traditional instructional methods. The form was designed using Microsoft Office 365 and distributed via email. This platform was selected because of its accessibility and user familiarity; all participants had access to Office 365 and were accustomed to using its features, ensuring ease of response submission. Data analysis We conducted a thematic analysis of open-ended questionnaire data from non-native Arabic learners using generative AI tools, allowing themes to emerge organically from their experiences (Creswell, 2014; Drisko & Maschi, 2016). Using NVivo 14, we followed Hsieh and Shannon’s (2005) three-phase procedure. First, we prepared the data—numbering each response, consolidating into a single file, and cleaning for coding. Second, we identified units of analysis (sentences or phrases reflecting learners’ perceptions), informed by AI-in-education research (Joo et al., 2016; Özgür, 2020), and developed an initial codebook from the literature and recurrent data patterns. A 10% pre-test by an external researcher refined the codes. In the third phase, two researchers independently applied the finalized codebook, achieving 86% intercoder agreement; disagreements were resolved through consensus (Creswell & Miller, 2000). This systematic, collaborative approach ensured credible identification of themes related to engagement, adaptivity, linguistic challenges, and perceived AI effectiveness. Results Research Question #1 How do non-native learners describe their experiences with learning Arabic using generative AI tools? The aim of the first research question was to explore learners’ experiences with studying Arabic as a non-native language and to understand the challenges they encountered throughout their learning journey. The 25 participants in this study—Greek-speaking adult learners residing across Europe and the Gulf region—shared diverse perspectives shaped by their linguistic backgrounds, learning environments, and personal motivations. Their responses revealed a rich blend of enthusiasm, cultural appreciation, and perseverance, but also highlighted persistent challenges such as mastering the Arabic script, adapting to unfamiliar grammar structures, and accessing high-quality instructional support. From their narratives, five major themes emerged: positive learning experiences, motivational factors, learning challenges, learning environments, and the early stages of acquisition. Each theme included several subthemes reflecting learners’ personal goals, emotional engagement, and the contextual barriers they face. Table 1 presents a synthesis of these themes, subthemes, and illustrative quotations drawn from participants’ lived experiences. Table 1 Themes and Subthemes for RQ1: Learners’ Experience with Learning Arabic Main Theme Subtheme Supporting Quotations Positive Learning Experience Enjoyment in Learning Arabic Participants highlighted that "My experience so far has been great, I enjoy learning languages." (S1) Cultural Appreciation Participants highlighted that " The Arabic language is interesting because of its symbols and its differences from other languages." (S2) Motivational Factors Professional Needs Participants highlighted that "I am working in Saudi Arabia, so I wanted to do this step for professional and personal reasons." (S12) Personal Interest Several learners emphasized "I am working in Saudi Arabia, so I wanted to do this step for professional and personal reasons." (S7) Challenges in Learning Alphabet and Script Learners frequently mentioned " "The learning pace was time-consuming, and the instruction was not oriented toward obtaining any certification." (S3) Lack of Certified Instruction Some responses revealed "It is a language that is unique, so you have to learn everything from zero." S6) Learning Environment Small Group Learning It was noted that "I am in a small group teaching course." (Participant s4) Independent Study It was noted that "I am in a small group teaching course." (S7) Early Learning Stage Limited Experience It was noted that "I’m just beginning my journey (two sessions in), so I don’t have all the full picture yet." (S8) Initial Enthusiasm A number of participants shared that "I’m just beginning my journey (two sessions in), so I don’t have all the full picture yet." (S12) Positive Learning Experience Some participants described their experience learning Arabic as highly enjoyable and personally fulfilling. For example, one learner shared, "My experience so far has been great, I enjoy learning languages." (S1). Others expressed a deep appreciation for the language’s cultural uniqueness, such as Participant S3 who noted, "The Arabic language is interesting because of its symbols and its differences from other languages." Motivational Factors Motivations for learning Arabic varied, but several participants emphasized both personal and professional goals. One participant, for instance, explained, "I am working in Saudi Arabia, so I wanted to do this step for professional and personal reasons." (S7). This reflects how learners are driven not only by career advancement but also by a genuine interest in engaging with the local culture and language. Challenges in Learning Despite their enthusiasm, learners reported notable challenges. A few participants found the Arabic alphabet and script especially difficult to master, with one commenting, "The Arabic language is interesting because of its symbols and its differences from other languages." (S2). Others noted the lack of structured or certified instruction as a barrier, as one participant described, "It is a language that is unique, so you have to learn everything from zero." (S9) Learning Environment Participants described varying learning environments, with some emphasizing the benefits of small group instruction. As one noted, "I am in a small group teaching course." (S4). Others reported a more self-directed learning experience, as reflected in another comment: "I am in a small group teaching course." (S11). While the contexts differed, these environments shaped how learners interacted with the language and the AI tools. Early Learning Stage Several participants were at the beginning of their Arabic learning journey and acknowledged their limited exposure so far. One learner remarked, "I’m just beginning my journey (two sessions in), so I don’t have all the full picture yet." (S5). This sense of initial enthusiasm combined with uncertainty was common among early-stage learners as they navigated the complexities of Arabic for the first time. Research Question 2 In what ways do generative AI tools provide personalized and adaptive support for Arabic language learners? The second research question aimed to examine how generative AI tools support personalized and adaptive learning experiences for non-native Arabic learners across varying proficiency levels. The participants described how tools such as Google Translate, ChatGPT, and Duolingo provided flexible, real-time assistance that catered to their individual learning needs. Their responses revealed six main themes related to adaptivity: personalized vocabulary support, pronunciation and listening features, self-paced learning and flexibility, grammar and sentence structure, learning confidence and control, and limitations in personalization. Within these themes, learners emphasized the usefulness of AI features such as synonym suggestions, audio pronunciation playback, and the ability to study at their own pace and on their own schedule. However, participants also noted shortcomings, particularly in handling Arabic dialects, delivering nuanced grammatical support, and providing culturally sensitive feedback. Table 2 summarizes the themes, subthemes, and participant quotations that reflect the multifaceted role of AI in delivering adaptive and learner-centered support in Arabic language acquisition. Table 2 Themes and Subthemes for RQ2: Adaptively and Personalized Support through AI Tools Main Theme Subtheme Supporting Quotations Personalized Vocabulary Support Word Suggestions Learners frequently mentioned "Google Translate proposes other words that have the same meaning, which helps deepen my Arabic knowledge." (S1) Synonym Alternatives Several learners emphasized " "It would form a sample sentence in English (S12) Visual Reinforcement Several learners emphasized "Duolingo helps a lot, especially for vocabulary." (S25) Pronunciation and Listening Features Audio Playback It was noted that "I have found AI really helpful regarding pronunciation." (Participant s1) Phonetic Transcription It was noted that " "It cannot recognize the dialect and suggests words that are far from the appropriate answer." (S17) Repetition Support Several learners emphasized "I use it to hear how the word sounds." (S21) Self-Paced Learning and Flexibility Independent Learning Participants highlighted that "There are gaps in support for the Greek language." (S19) Learning Anytime Participants highlighted that "AI allows fast access to information and boosts productivity." (S23) Pacing According to Need Several learners emphasized "I think if you use it for specific words it can be helpful to move on in an exercise." (S22) Grammar and Sentence Structure Root Word Discovery Participants highlighted that "I usually use it to find the root of the verbs." (S22) Grammar Assistance A number of participants shared that "I used it to form a sentence as an example in the English language. (S16)) Sentence Examples Several learners emphasized "ChatGPT is used mostly for document translation in my job." (S5) Limitations in Personalization Dialect Recognition Issues A number of participants shared that “one of its limitations is It cannot recognize the dialect and suggests words that deviate from the accurate answer (S11) Low Support for Greek Some responses revealed "There are deficiencies in the Greek language support" (S14) Generalized Feedback Participants highlighted that "Sometimes when translating a sentence, the meaning makes no sense." (S3) Learning Confidence and Control Assurance When Stuck Learners frequently mentioned "It doesn’t make me feel more confident, but it makes me feel secure if I get stuck." (S10) Reduced Fear of Mistakes Participants highlighted that " For me, there is no fear of failure of using it [AI tools] (S6) Confidence Building Over Time It was noted that "Generative AI can be helpful as a tool in general for all those fields." (S12) Personalized Vocabulary Support Many participants appreciated the way AI tools enhanced their vocabulary development. For example, one learner explained, "Google Translate proposes other words that have the same meaning, which helps deepen my Arabic knowledge." (S1). Another participant noted how AI tools provided semantic alternatives and explanations, stating, "It often presented words with related meanings and explained the meaning of the word in a descriptive way." (S5). In addition, several learners pointed to the effectiveness of Duolingo in reinforcing vocabulary through repetition and visuals, with one sharing, "Duolingo helps a lot, especially for vocabulary." (S11) Pronunciation and Listening Features Participants emphasized the usefulness of AI in supporting pronunciation and listening comprehension. One learner shared, "I have found AI really helpful regarding pronunciation." (S14), while another highlighted the benefit of auditory repetition: "These tools also have the option for the learner to hear the word." (S18). Similarly, one participant explained, "I use it to hear how the word sounds." (S17), indicating how repetition can enhance phonetic awareness. Self-Paced Learning and Flexibility Several participants emphasized the flexibility AI tools offer for independent and self-directed learning. For instance, one noted, "It [AI tools] promotes equal opportunities and the learner’s self-initiative while learning new languages." (S21). Others appreciated the immediate access to information, as expressed by Participant S25: "AI allows fast access to information and boosts productivity." Additionally, some learners reported that AI helped them progress at their own pace, with one stating, "I think if you use it for specific words it can be helpful to move on in an exercise." (S13) Grammar and Sentence Structure Participants also described how AI tools supported their understanding of Arabic grammar and sentence construction. One learner reflected, "It helped me try to identify the root forms of verbs, which supported my understanding of how the language is structured." (S23). Another noted that AI tools provided examples in English that aided comprehension, saying, "It would form a sample sentence in English." (S12). In workplace contexts, some participants reported using AI primarily for functional translation tasks, as one stated, "ChatGPT is used mostly for document translation in my job." (S15) Limitations in Personalization Despite the benefits, participants also highlighted key limitations in AI personalization. Several learners noted that dialect recognition remains a major challenge, with one stating, "It cannot recognize the dialect and suggests words that are far from the appropriate answer." (S17). Others pointed to insufficient language support, particularly in Greek, explaining, "There are gaps in support for the Greek language." (S19). A few participants also expressed frustration with generalized or inaccurate feedback, such as: "Sometimes when translating a sentence, the meaning makes no sense." (S21) Learning Confidence and Control Participants reported mixed experiences regarding how AI tools influenced their confidence in language use. One participant expressed a sense of security rather than confidence, stating, "It doesn’t make me feel more confident, but it makes me feel secure if I get stuck." (S6). Others found reassurance in the learning process, such as one who noted, "There is no fear of failure." (S3). Additionally, some participants recognized AI’s potential as a supportive tool over time, with one explaining, "Generative AI can be helpful as a tool in general for all those fields." (S4) Research Question #3: What challenges do learners face when using generative AI tools for Arabic language learning? The third research question focused on identifying the challenges and limitations that non-native learners encounter when using generative AI tools to support their Arabic language learning. Participants reported a variety of concerns, which we organized into five key themes: translation inaccuracy, dialect and cultural nuance limitations, dependence and reduced critical thinking, limited customization, and language support inequities. Learners frequently noted that AI-generated translations often failed to capture the intended meaning, especially when translating full sentences or culturally rich expressions. Others expressed frustration with the tools’ inability to differentiate between dialects or provide accurate context-based feedback. Additionally, several participants reported a growing dependence on AI for tasks such as spelling or grammar correction, which led to reduced memory recall and critical language processing. Concerns were also raised about the limited support for the Greek language and the dominance of English in AI responses. Table 3 presents these themes and subthemes along with selected participant quotations that illustrate the barriers encountered in using AI for Arabic language learning. Table 3 Themes and Subthemes for RQ3: Challenges and Limitations of Using AI Tools in Arabic Learning Main Theme Subtheme Supporting Quotations Translation Inaccuracy Meaning Distortion It was noted that "Most of the times I can’t sufficiently pass the message that I intend." (S7) Grammar Errors It was noted that "Sometimes when you try to translate a sentence, the meaning will make no sense." (S13) Lack of Context Several learners emphasized "I’ve noticed that translations into Greek are not always accurate, which can make it harder for me to fully understand the intended meaning." (S12) Dialect and Cultural Nuance Limitations Dialect Misinterpretation Learners frequently mentioned "It struggles to translate proverbs, idiomatic expressions, and dialects accurately, which affects my ability to fully understand cultural and contextual meanings." (S18) Idiom Misunderstanding Several learners emphasized "Minor issues exist that an expert translator should finalize." (S9) Cultural Gaps It was noted that "You can’t trust it 100%." (S9) Dependence and Reduced Critical Thinking Overreliance on AI It was noted that " I often rely on it to check the spelling of words, which makes me less likely to retain them independently." (S23) Memory Inhibition It was noted I don’t actively engage my thinking and memory, which leads to noticeable gaps in my learning over time. (S17) Copy-Paste Learning Learners frequently mentioned "It’s helpful, but it doesn’t build confidence." (S6) Limited Customization Lack of Personalized Feedback Participants highlighted that "I would like to see more tools focused in specific fields like Geography, science etc." (S4) Insufficient Task Variety Some responses revealed " "It would be beneficial if learners had the option to create their own quizzes and receive clear explanations for any mistakes.(S3) Topic Irrelevance Some responses revealed ""it could support classroom learning by providing flashcards with images." (S10) Language Support Inequities Stronger English Support Some responses revealed " there seem to be more features available in English." (S15) Weaker Greek Integration Participants highlighted "I don’t really trust the information given by AI." (S2) Tool Limitations Some responses revealed "I don’t really trust the information given by AI." (Participant s1) Translation Inaccuracy A number of participants expressed concerns about the limitations of AI-generated translations in conveying accurate meaning. For instance, one learner remarked, "Most of the times I can’t sufficiently pass the message that I intend." (S1). Several participants also emphasized issues related to grammatical accuracy, with one stating, "Sometimes when you try to translate a sentence, the meaning will make no sense." (S6). Additionally, some participants noted that poor contextual understanding often hindered comprehension, as reflected in the comment: "The translation into Greek is not always accurate, which sometimes affects my understanding of the intended meaning." (S8) Dialect and Cultural Nuance Limitations Some participants highlighted the inability of AI tools to handle the complexity of dialects and culturally embedded expressions. One participant observed, "It is unable to accurately translate proverbs, idiomatic expressions, and dialects, which limits my ability to grasp the deeper meaning of certain phrases." (S9). Others noted that outputs often lacked cultural refinement, with one learner stating, "Minor issues exist that an expert translator should finalize." (S4). A few participants also expressed general distrust, emphasizing that "You can’t trust it 100%." (S15) Dependence and Reduced Critical Thinking Several participants raised concerns about becoming overly reliant on AI tools for routine language tasks. One learner admitted, "I often rely on AI to check the spelling of words, which makes me less likely to recall them on my own." (S18). Others reflected on how this dependence affects cognitive engagement, such as Participant Sa4, who noted, "I don’t actively engage my thinking and memory skills, which sometimes results in gaps in my language use." Additionally, some learners questioned the long-term benefits of AI-supported learning, with one stating, "It’s helpful, but it doesn’t build confidence." (S5). Limited customization The majority of participants emphasized the need for domain-specific resources that align with learners’ academic or professional interests. For example, one participant noted, "I would like to see more tools focused in specific fields like Geography, science, etc." (S4). Moreover, a few participants highlighted the importance of interactive features that promote learner autonomy, expressing a desire for "the option to create their own quizzes and receive explanations for their mistakes" (S16). On the other hand, some participants pointed to the value of visual aids in supporting vocabulary development, suggesting that "it could support classroom learning by providing visual flashcards to help with vocabulary acquisition." (S19). Language support inequities A number of participants observed that generative AI tools tend to offer more robust functionality in English compared to other languages. As one participant noted, "I notice that there are more features available in English." (S11). In addition, several participants pointed out limited integration and support for the Greek language, with one learner stating, "There are gaps in support for the Greek language." (S7). Some participants also expressed concerns about the reliability of AI-generated content, as reflected in the comment: "I don’t really trust the information given by AI." (S2). Discussion Learners in this study consistently reported that generative AI tools transformed their Arabic studies from a series of isolated drills into an engaging, culturally rich experience. Where traditional gamified apps rely on points or badges to motivate, AI-powered conversational agents simulated realistic dialogues that reduced anxiety and encouraged experimentation with new vocabulary and structures (Almelhes, 2024 ; Kotob & Ibrahim, 2019 ). These interactions often felt more meaningful than rote practice, since learners could immediately see how words and phrases functioned in context. In doing so, AI did not simply supplement instruction; it created its own micro-environments in which emotional engagement with the language flourished, echoing early findings on the low‐stakes affordances of chatbots (Fryer & Carpenter, 2006 ). The adaptive feedback mechanisms built into these systems played a crucial role in scaffolding learner progress. Unlike one-size‐fits‐all exercises, AI continuously analyzed individual performance—suggesting synonyms, adjusting pronunciation models, and calibrating task difficulty to each user’s proficiency (Putri et al., 2021; Nasaruddin, 2024 ). This immediacy enabled learners to target specific weaknesses and to consolidate gains through spaced, iterative practice that mirrored the personalized pathways of intelligent tutoring systems (Putri et al., 2021). In effect, AI became a virtual tutor capable of tailoring instruction in real time, fostering self‐regulated study habits that many learners found both efficient and empowering. Despite these advances, significant gaps emerged around translation accuracy and cultural nuance. Participants frequently encountered misrendered idioms or literal translations that failed to convey intended meaning—limitations long documented in natural language processing research (Coniam, 2014 ; Zhao & Zhang, 2019 ). AI’s difficulty with regional dialects further underscored the challenge of modeling Arabic’s sociolinguistic diversity (Dokukina & Gumanova, 2020 ). When tools produced outputs that felt “off” or culturally insensitive, learners had to rely on external resources or instructor guidance to resolve misunderstandings. These persistent shortcomings highlight the need for richer, more dialect-aware corpora and for collaborative refinement between AI developers and language experts. A subtler concern centered on cognitive overreliance. While many learners appreciated that AI provided a safety net for spelling, grammar, or pronunciation, several noted that habitual use sometimes weakened their own recall and analytical engagement. This phenomenon has been identified in reviews of chatbot-supported learning, which caution that easy access to answers can inadvertently diminish active problem‐solving and memory consolidation (Huang, Hew, & Fryer, 2022 ). Without deliberate instructional design to counterbalance AI assistance—such as tasks requiring learners to justify or critique AI suggestions—there is a risk that automated feedback might supplant rather than support deeper processing. Importantly, these challenges did not diminish AI’s overall value but rather pointed to the necessity of a blended, multimodal ecosystem. Comparative studies have shown that no single tool sufficiently addresses every pedagogical need (Stošić & Guillén-Gámez, 2024; Dokukina & Gumanova, 2020 ). In our context, the richest learning emerged when AI‐driven drills were integrated with immersive, human‐mediated activities: live, instructor‐led discussions that contextualized AI practice within broader cultural and communicative frameworks. This combination allowed learners to leverage AI’s adaptability while still benefiting from the nuanced corrections and sociocultural insights that only a teacher or native speaker can provide. Ensuring equitable access further emerged as a critical condition for success. High-speed internet, up‐to‐date hardware, and reliable platforms are prerequisites for seamless AI interactions—resources that are unevenly distributed across regions and institutions (Azmi & Zakaria, 2020 ; Wulantari et al., 2023 ). Without robust infrastructure investments and clear data‐privacy policies, the promise of AI in language education risks exacerbating existing inequities. Moreover, educators themselves require ongoing professional development to navigate AI tools effectively, interpret their outputs, and integrate them into pedagogical strategies that preserve learner autonomy and critical thinking (Pokrivčáková, 2019). Theoretical and practical implications For Researchers: This study advances AI-in-language-education research by using Arabic—a linguistically and culturally complex case—to evaluate generative AI tools’ effectiveness in delivering personalized learning, adaptive feedback, and sustained engagement across diverse learners. It calls for interdisciplinary frameworks uniting pedagogy, AI, and sociolinguistics. Future work should examine long-term outcomes, learner autonomy, and cognitive engagement, while addressing AI’s limitations with underrepresented dialects. Scholars can also improve AI models’ accuracy in translation, cultural nuance detection, and pronunciation support through rigorous evaluation and iterative refinement. For Students: From a theoretical perspective, our findings highlight generative AI’s capacity to scaffold autonomous learning, accelerate vocabulary development, and bolster both oral and written Arabic proficiency—especially by adapting to complex phonetic and grammatical challenges. Practically, tools like ChatGPT and Duolingo enable learners to reinforce classroom instruction, obtain instant clarification, and gain linguistic confidence. Yet, effective use requires digital literacy: students must critically evaluate AI-generated material and guard against overreliance. Ultimately, AI should serve as a complementary aid—enhancing human-led instruction and cultural immersion rather than replacing them. For Teachers: Theoretically, this study underscores educators’ pivotal role in mediating AI-driven language learning. Rather than displacing teachers, AI enhances their function as facilitators who interpret system outputs, guide learner reflection, and offer culturally grounded feedback. Practically, it advocates for teacher training that blends digital pedagogy with AI literacy. Educators should learn to embed AI tools in lesson design, tailor instruction using AI-generated diagnostics, and coach students on ethical AI engagement. Blended models—pairing generative AI with live teaching—can yield more adaptive, learner-centered classrooms. For Decision Makers: This study argues that AI should be integrated into education policy as part of a comprehensive digital ecosystem rather than treated in isolation. Language planners and institutional leaders must adopt evidence-based approaches to embed AI in curricula and assessment inclusively. Practically, they should invest in infrastructure for equitable AI access—especially in non-native Arabic contexts—and support AI development that respects linguistic and cultural diversity. Continuous professional development for teachers, alongside clear ethical guidelines and quality standards, is essential to harness AI’s benefits while mitigating risks of bias, privacy breaches, and misinformation. Limitations and Future Research Limitations While offering valuable insights into generative AI’s support for non-native Arabic learners, this study has several limitations. First, it relies on self-reported data from only 25 participants, which may restrict generalizability and reflect individual biases, expectations, and digital-literacy levels. Second, its focus on Greek-speaking learners—though culturally varied—does not capture the full spectrum of global Arabic students. Third, it addresses only Modern Standard Arabic, overlooking regional dialects prevalent in everyday communication. Finally, it omits teachers’ and AI developers’ perspectives, which could have deepened understanding of pedagogical and technological design implications. Future Research Future research should broaden participant samples to encompass greater linguistic and cultural diversity, enabling comparative analyses of how AI supports Arabic learning across different native-language groups. Studies on AI integration for dialectal Arabic—addressing regional variants, idioms, and sociolinguistic nuances—are also needed. Longitudinal designs would illuminate the sustained effects of generative AI on proficiency and motivation. Including teachers’ and instructional designers’ perspectives can yield a more comprehensive understanding of AI’s curricular integration. Finally, experimental comparisons of AI-supported, blended, and traditional methods will empirically validate the pedagogical value of these technologies in real-world settings. Conclusion This article has examined how generative AI transforms Arabic language learning for non-native speakers by shifting pedagogy from rote vocabulary drills to immersive, voice-centric practice. Leveraging platforms such as ChatGPT, Duolingo, and Google Translate, these AI systems create culturally authentic dialogue simulations that reduce learner anxiety and foster sustained engagement. Through real-time, adaptive feedback—spanning spaced-repetition vocabulary reinforcement, dynamic pronunciation modeling, and context-sensitive grammar scaffolding—AI tailors instruction to each learner’s evolving needs. Performance analytics drive automatic adjustments in task difficulty and pacing, fostering self-regulated study and individualized learning pathways. Yet, persistent limitations—chiefly dialect misrecognition, occasional semantic inaccuracies, and the risk of overreliance—underscore that AI cannot fully replicate the nuanced insights of human instruction. The most effective implementations therefore blend AI-driven exercises with instructor-mediated dialogue, ensuring culturally grounded correction, critical reflection, and personalized cultural context. As AI architectures integrate richer dialectal corpora and more sophisticated neural models, they promise to further narrow the gap between vocabulary acquisition and vocal fluency. Ultimately, the strategic fusion of generative AI’s scalable adaptability with pedagogical expertise offers a scientifically grounded, culturally responsive framework for achieving meaningful spoken mastery of Modern Standard Arabic and its regional variants. Declarations Ethics Approval and Consent to Participate This study was reviewed and approved by the Institutional Review Board (IRB) at XXX (Approval Number: Lang. Feb. 2025/13). All procedures involving human participants were conducted in accordance with institutional guidelines and the ethical standards of the American Psychological Association (APA, 2010) and the Declaration of Helsinki (2013). Informed consent was obtained from all participants prior to participation. Participants were informed of the study’s purpose, the voluntary nature of their involvement, and their right to withdraw at any time without penalty. Consent was obtained verbally and documented, as approved by the IRB. Human Ethics and Consent to Participate Declarations All relevant human ethics and consent procedures were followed. There are no additional declarations beyond those described above. Funding This research received no specific grant or financial support from any funding agency, commercial, or not-for-profit sectors. Author Contribution A.K., B.H., A.K.E., S.N., and Z.N.K. contributed to the conception and design of the study. A.K., B.H., A.K.E., and S.N. contributed to data collection and preliminary analysis. Z.N.K. supervised the research process and contributed to the interpretation of the findings. A.K. and Z.N.K. wrote the main manuscript text. All authors reviewed, revised, and approved the final manuscript. Data Availability Data will be available upon request from the corresponding author References Al-Bulushi, A. H., & Al‐Issa, A. S. (2017). Playing with the language: Investigating the role of communicative games in an Arab language teaching system. International Journal of Instruction , 10 , 179–198. https://doi.org/10.12973/iji.2017.10212a Alenezi, M. A. K., Mohamed, A. M., & Shaaban, T. S. (2023). Revolutionizing EFL special education: How ChatGPT is transforming the way teachers approach language learning. International Journal of Technology and Educational Innovation , 9 (2), 5–23. Almelhes, S. A. (2023). A review of artificial intelligence adoption in second-language learning. Theory and Practice in Language Studies , 13 (5), 1259–1269. https://doi.org/10.17507/tpls.1305.21 Almelhes, S. A. (2024). Gamification for teaching the Arabic language to non-native speakers: A systematic literature review. Frontiers in Education , 9 ., Article 1371955. https://doi.org/10.3389/feduc.2024.1371955 Azmi, N. H., & Zakaria, Z. M. (2020). Gamification among UPSI training teacher during teaching practice in school. International Journal of Modern Education , 2 , 56–67. https://doi.org/10.35631/IJMOE.24005 Becker, S. A., Brown, M., Dahlstrom, E., Davis, A., DePaul, K., Diaz, V., & Pomerantz, J. (2018). NMC horizon report: 2018 higher education edition. Educause. Retrieved from https://cit.bnu.edu.cn/docs/2018-09/20180918163624337480.pdf Bjelajac, Ž., Filipović, A., & Stošić, L. (2022). Quis custodiet ipsos custodes: Ethical dilemmas of the KM governed by AI. In M. Stanković & V. Nikolić (Eds.), Proceedings of the 4th Virtual International Conference Path to a Knowledge Society: Managing Risks and Innovation (pp. 353–360). SASA. Retrieved from https://miteam.mi.sanu.ac.rs/asset/G3sjTM6rPPZThzpnM Bjelajac, Ž., Filipović, A. M., & Stošić, L. (2023). Can AI be evil: The criminal capacities of ANI. International Journal of Cognitive Research in Science Engineering and Education , 11 (3), 519–531. https://doi.org/10.23947/2334-8496-2023-11-3-519-531 Burgsteiner, H., Kandlhofer, M., & Steinbauer, G. (2016, March). Irobot: Teaching the basics of artificial intelligence in high schools. Proceedings of the AAAI Conference on Artificial Intelligence, 30 (1). https://doi.org/10.1609/aaai.v30i1.9864 Coniam, D. (2014). The linguistic accuracy of chatbots: Usability from an ESL perspective. Text & Talk , 34 (5), 545–567. https://doi.org/10.1515/text-2014-0018 Dokukina, I., & Gumanova, J. (2020). The rise of chatbots: New personal assistants in foreign language learning. Procedia Computer Science , 169 , 542–546. Fathi, J., Rahimi, M., & Derakhshan, A. (2024). Improving EFL learners’ speaking skills and willingness to communicate via artificial intelligence-mediated interactions. System , 121 . https://doi.org/10.1016/j.system.2024.103254 . Article 103254. Fryer, L. K., & Carpenter, R. (2006). Bots as language learning tools. Language Learning & Technology , 10 (3), 8–14. https://dx.doi.org/10.125/44068 García Botero, G., Questier, F., & Zhu, C. (2019). Self-directed language learning in a mobile-assisted, out-of-class context: Do students walk the talk? Computer Assisted Language Learning , 32 (1–2), 71–97. https://doi.org/10.1080/09588221.2018.1485707 Gafni, R., Achituv, D. B., & Rachmani, G. J. (2017). Learning foreign languages using mobile applications. Journal of Information Technology Education: Research , 16 , 301–317. Grishechko, E. G. (2021). Communicative strategy of politeness in modern English: Multi-level means of implementation . EHERPQ. Grishechko, E. G. (2023, April 7). Enhancing language learning with podcasts: Evidence from feedback and practice. In Proceedings of the 2nd Pan-Russian Research and Practical Conference on Topical Issues of Intercultural Communication and Foreign Language Teaching Methodology (pp. 69–72). AkademLit. Grishechko, E. G. (2024). Analysis of the potential of the Fusion Brain neural network as a tool for visualizing training materials. In Proceedings of the 11th International Research and Practical Conference on Topical Issues of Linguistics and Teaching Methods in Business and Professional Communication (pp. 345–347). RUDN University. Guillén-Gámez, F. D., Lugones, A., & Mayorga-Fernández, M. J. (2019). ICT use by pre-service foreign languages teachers according to gender, age and motivation. Cogent Education , 6 (1). Article 1574693. https://doi.org/10.1080/2331186X.2019.1574693 Hong, W. C. H. (2023). The impact of ChatGPT on foreign language teaching and learning: Opportunities in education and research. Journal of Educational Technology and Innovation, 5 (1). Retrieved from https://jeti.thewsu.org/index.php/cieti/article/view/103/64 Huang, W., Hew, K. F., & Fryer, L. K. (2022). Chatbots for language learning: Are they really useful? A systematic review of chatbot-supported language learning. Journal of Computer Assisted Learning , 38 (1), 237–257. https://doi.org/10.1111/jcal.12610 Kasneci, E., Seßler, K., Küchemann, S., Bannert, M., Dementieva, D., Fischer, F., & Kasneci, G. (2023). ChatGPT for good? On opportunities and challenges of large language models for education. Learning and Individual Differences , 103 , 102274. https://doi.org/10.1016/j.lindif.2023.102274 Klimova, B. (2020). Benefits of the use of mobile applications for learning a foreign language by elderly population. Procedia Computer Science , 176 , 2184–2191. https://doi.org/10.1016/j.procs.2020.09.255 Klimova, B., & Pikhart, M. (2019). Cognitive and applied linguistics aspects of using social media: The impact of the use of Facebook on developing writing skills in learning English as a foreign language. European Journal of Investigation in Health Psychology and Education , 10 (1), 110–118. https://doi.org/10.3390/ejihpe10010010 Kong, S. C., Cheung, W. M. Y., & Zhang, G. (2021). Evaluation of an artificial intelligence literacy course for university students with diverse study backgrounds. Computers and Education: Artificial Intelligence , 2 , 100026. https://doi.org/10.1016/j.caeai.2021.100026 Kenali, H. M. S., Yusoff, N. M. R. N., Saad, M., Abdullah, N. S. B., H., & Kenali, A. M. S. (2019). The effects of language games on smartphones in developing Arabic speaking skills among non-native speakers. Creative Education , 10 , 972–979. https://doi.org/10.4236/ce.2019.105073 Kotob, M. M., & Ibrahim, A. (2019). Gamification: The effect on students’ motivation and achievement in language learning. Journal of Applied Linguistics & Language Research , 6 , 177–198. https://doi.org/10.4995/eurocall.2020.12974 Lazar, S., & Guillén-Gámez, F. D. (2024). The potential of IT tools in foreign language acquisition: A comparative assessment. Training Language and Culture , 8 (4), 95–108. https://doi.org/10.22363/2521-442X-2024-8-4-95-108 Loewen, S., Crowther, D., Isbell, D. R., Kim, K. M., Maloney, J., Miller, Z. F., & Rawal, H. (2019). Mobile-assisted language learning: A Duolingo case study. ReCALL , 31 (3), 293–311. https://doi.org/10.1017/S0958344019000340 Lord, G. (2015). I don't know how to use words in Spanish’: Rosetta Stone and learner proficiency outcomes. The Modern Language Journal , 99 (2), 401–405. https://doi.org/10.1111/modl.12234_3 Lu, X. (2019). An empirical study on the artificial intelligence writing evaluation system in China CET. Big Data , 7 (2), 121–129. https://doi.org/10.1089/big.2018.0151 Malyuga, E. N. (2007). Functional-pragmatic features of speech behavior of British and American participants in intercultural business communication . RUDN University. Malyuga, E. N. (2008). A newspaper text: Genre and style paradigm in English interrogative sentences. Bulletin of the University of the Russian Academy of Education , 2 , 68–71. Malyuga, E. N. (2010). Professionally oriented foreign language teaching at the present stage. Issues of Applied Linguistics , 3 , 83–97. Malyuga, E. N. (2019). Emergent trends in English scientific discourse: Issues of research relevance and linguistic identity. Tomsk State University Journal of Philology , 58 , 52–70. https://doi.org/10.17223/19986645/58/4 Mellati, M., Khademi, M., & Abolhassani, M. (2018). Creative interaction in social networks: Multi-synchronous language learning environments. Education and Information Technologies , 23 , 2053–2071. https://doi.org/10.1007/s10639-018-9703-9 Mohamed, A. M., Shaaban, T. S., Bakry, S. H., Guillén-Gámez, F. D., & Strzelecki, A. (2024). Empowering the faculty of education students: Applying AI’s potential for motivating and enhancing learning. Innovative Higher Education . https://doi.org/10.1007/s10755-024-09747-z Nasaruddin, N. (2024). Using ChatGPT in teaching Arabic as a foreign language. Arabiyatuna: Jurnal Bahasa Arab , 8 (1), 1–24. http://dx.doi.org/10.29240/jba.v8i1.9413 Ng, D. T. K., Leung, J. K. L., Chu, S. K. W., & Qiao, M. S. (2021). Conceptualizing AI literacy: An exploratory review. Computers and Education: Artificial Intelligence , 2 , 100041. https://doi.org/10.1016/j.caeai.2021.100041 Nguyen, T. X. (2024). English majors’ perceptions of AI tool application in English language learning at tertiary level in Vietnam. Journal of Knowledge Learning and Science Technology , 3 (1), 179–193. https://doi.org/10.60087/jklst.vol3.n1.p193 Panagiotidis, P. (2021). Augmented and mixed reality in language learning. European Journal of Education , 4 (2), 27–43. https://doi.org/10.26417/501ibq23c Petrosyan, G. O., & Grishechko, E. G. (2020). Integration of linguistic and didactic approaches in the process of foreign language mastering. Modern Humanities Success , 11 , 125–131. Pratama, M. P., Sampelolo, R., & Lura, H. (2023). Revolutionizing education: Harnessing the power of artificial intelligence for personalized learning. Klasikal: Journal of Education Language Teaching and Science , 5 (2), 350–357. https://doi.org/10.52208/klasikal.v5i2.877 Rouffet, C., van Beuningen, C., & de Graaff, R. (2023). Constructive alignment in foreign language curricula: An exploration of teaching and assessment practices in Dutch secondary education. The Language Learning Journal , 51 (3), 344–358. https://doi.org/10.1080/09571736.2022.2025542 Şahín Kölemen, C. (2024). Artificial intelligence technologies and ethics in educational processes: Solution suggestions and results. Innoeduca: International Journal of Technology and Educational Innovation , 10 (2), 201–216. https://doi.org/10.24310/ijtei.102.2024.19806 Salma, N. (2020). Collaborative learning: An effective approach to promote language development. International Journal of Social Sciences & Educational Studies , 7 (2), 57–61. https://doi.org/10.23918/ijsses.v7i2p57 Sohail, S. S., Farhat, F., Himeur, Y., Nadeem, M., Madsen, D. Ø., Singh, Y., & Mansoor, W. (2023). Decoding ChatGPT: A taxonomy of existing research, current challenges, and possible future directions. Journal of King Saud University – Computer and Information Sciences , 25 (8). Article 101675. https://doi.org/10.1016/j.jksuci.2023.101675 Stošić, L., & Janković, A. (2023). The impact of artificial intelligence (AI) on education: Balancing advancements and ethical considerations on human rights. Pravo – Teorija i Praksa , 40 (4), 58–72. https://doi.org/10.5937/ptp2304058S Stošić, L., & Mikhailova, O. B. (2023). Leadership and innovativeness as the basis for developing teachers’ digital competences. RUDN Journal of Psychology and Pedagogics , 20 (1), 126–144. https://doi.org/10.22363/2313-1683-2023-20-1-126-144 Talan, T., Doğan, Y., & Kalinkara, Y. (2024). Digital natives’ mobile learning adoption in terms of UTAUT-2 Model: A structural equation model. Innoeduca: International Journal of Technology and Educational Innovation , 10 (1), 100–123. https://doi.org/10.24310/ijtei.101.2024.17440 Valencia, E. M., Rivas, E. S., Palmero, J. R., & Gámez, F. D. G. (2022). Perceptions of university students about virtual reality as a didactic resource: A pre-experimental study with a control and experimental group. IJERI: International Journal of Educational Research and Innovation , 17 , 152–171. https://doi.org/10.46661/ijeri.5904 Viktoria, D., Polina, L., Natalia, A., Lilia, N., & Evgenia, E. (2018, September). Virtual and augmented reality in language acquisition. In Proceedings of the International Conference on the Theory and Practice of Personality Formation in Modern Society (ICTPPFMS 2018) (pp. 218–223). Atlantis Press. https://doi.org/10.2991/ictppfms-18.2018.38 Woo, J. H., & Choi, H. (2021). Systematic review for AI-based language learning tools. ArXiv, 2111.04455 . https://doi.org/10.48550/arXiv.2111.04455 Wulantari, N. P., Rachman, A., Sari, M. N., Uktolseja, L. J., & Rofi’i, A. (2023). The role of gamification in English language teaching: A literature review. Journal on Education , 6 (1), 2847–2856. https://doi.org/10.31004/joe.v6i1.3328 Zhang, X., Romero-Forteza, F., & Li, Z. (2023). Bibliometric analysis of technology-supported language learning: LMOOC trend in China. Innoeduca: International Journal of Technology and Educational Innovation , 9 (2), 68–81. https://doi.org/10.24310/innoeduca.2023.v9i2.17362 Zhao, X., & Zhang, R. (2019). A state-of-the-art review of automated writing evaluation systems for language learning. Computer Assisted Language Learning , 35 (9), 2790–2816. https://doi.org/10.1080/09588221.2021.1896555 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9249983","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":617414548,"identity":"c0ec8ca5-59da-49e4-be49-d1765543139b","order_by":0,"name":"Ahmed Khouli","email":"","orcid":"","institution":"An-Najah National University","correspondingAuthor":false,"prefix":"","firstName":"Ahmed","middleName":"","lastName":"Khouli","suffix":""},{"id":617414549,"identity":"f262124f-1206-4587-9090-d90fe234ff87","order_by":1,"name":"Abed Alkhaleq Esa","email":"","orcid":"","institution":"An-Najah National University","correspondingAuthor":false,"prefix":"","firstName":"Abed","middleName":"Alkhaleq","lastName":"Esa","suffix":""},{"id":617414550,"identity":"0d2e0863-a600-4d03-9623-2ee941a442e5","order_by":2,"name":"Bilal Hamamra","email":"","orcid":"","institution":"An-Najah National University","correspondingAuthor":false,"prefix":"","firstName":"Bilal","middleName":"","lastName":"Hamamra","suffix":""},{"id":617414551,"identity":"0bab00c9-712a-462f-ab6d-e6ac3d3dff95","order_by":3,"name":"Suhad Nierat","email":"","orcid":"","institution":"An-Najah National University","correspondingAuthor":false,"prefix":"","firstName":"Suhad","middleName":"","lastName":"Nierat","suffix":""},{"id":617414552,"identity":"4c75fe8a-cd08-4f59-aec0-116e47f2eea3","order_by":4,"name":"Zuheir N Khlaif","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA0UlEQVRIie3OMQrCMBTG8VcEXZ64CtX2Cp9kVDxLpFCXFhwdPUB01lt0E7eUroqr4KK7m4tgQdPNLXUTzH8IGd4veUQu1y/G7GlCEJirbJjDW9QgZIgQ3xISk2qyHgnVQevHDNNtS1/vMxr1M20h3nIlcwWkOyUjf02xsJJGh6EZZZppKX2mYmIlTUPyEpjieImeTC874bZCwYDEScbmF20nXd6j6AGD7HSJh4xIbGwkVMngeisR4phEZ56P+ysb+dxREqH+eFXri+ddLpfrr3oDJqlDmwsR2isAAAAASUVORK5CYII=","orcid":"","institution":"An-Najah National University","correspondingAuthor":true,"prefix":"","firstName":"Zuheir","middleName":"N","lastName":"Khlaif","suffix":""}],"badges":[],"createdAt":"2026-03-28 05:53:44","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9249983/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9249983/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":109228656,"identity":"6c0fc037-2860-40aa-bf42-1c6a4a87209c","added_by":"auto","created_at":"2026-05-14 02:10:09","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":362110,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9249983/v1/5a5e5f0b-6bb9-4230-8cec-550b03b73991.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"From Vocabulary to Voice: How Generative AI Shapes Arabic Learning for Non-Native Speakers","fulltext":[{"header":"Introduction","content":"\u003cp\u003eIn recent years, artificial intelligence (AI) has emerged as a transformative force in language education, offering unprecedented opportunities to personalize, adapt, and enhance the learning experience for students across diverse linguistic and cultural contexts. AI-powered tools such as intelligent tutoring systems (ITS), natural language processing (NLP), automated writing evaluation (AWE), and chatbots are increasingly used in second and foreign language instruction to support engagement, adaptivity, and linguistic proficiency (Son, Ružić, \u0026amp; Philpott, 2023). These technologies provide learners with real-time feedback, tailored content, and immersive environments that simulate authentic communicative situations, thereby addressing many of the limitations of traditional, one-size-fits-all instruction.\u003c/p\u003e \u003cp\u003eThe potential of AI in Arabic language education has gained particular attention due to the language\u0026rsquo;s inherent complexity, including its rich morphology, diglossia, and challenging phonetic features (AlAfnan, 2025). AI-driven tools are especially useful for supporting non-native speakers in distinguishing between similar Arabic sounds, improving intonation, and developing more accurate pronunciation models\u0026mdash;tasks often difficult to achieve through conventional instruction alone. Moreover, adaptive AI systems can respond to individual learners\u0026rsquo; progress by modifying content and pacing, thereby creating a dynamic and inclusive learning experience (Pokrivč\u0026aacute;kov\u0026aacute;, 2019).\u003c/p\u003e \u003cp\u003eDespite this potential, the application of AI in teaching Arabic to non-native speakers remains underexplored compared to its use in English or other widely taught languages. While studies have emphasized the transformative impact of AI on Arabic language acquisition for native speakers (Alsaied, 2024), research focused on adult learners from diverse cultural and professional backgrounds, such as Greek-speaking students learning Arabic across multiple countries, is still limited. These learners, with motivations ranging from professional development to intercultural communication, often face additional challenges such as dialect variation, formal-informal language distinctions, and pronunciation accuracy\u0026mdash;all areas where AI can provide meaningful support (Son et al., 2023; Pokrivč\u0026aacute;kov\u0026aacute;, 2019).\u003c/p\u003e \u003cp\u003eFurthermore, scholars have called for a more integrated and theory-driven exploration of how AI can be implemented in foreign language education, not only to support technical aspects of learning but also to foster emotional engagement and cultural sensitivity (AlAfnan, 2025). This is especially relevant in Arabic instruction, where learners benefit from emotional connection and context-based use of the language.\u003c/p\u003e \u003cp\u003eAgainst this backdrop, the present study investigates how AI tools contribute to enhancing engagement, adaptively, and proficiency in Arabic among non-native speakers. Focusing on a diverse cohort of Greek-speaking learners located in the Gulf region, Greece, and other European countries, this research aims to provide empirical insights into how AI technologies mediate the experience of learning Arabic in virtual, culturally diverse settings.\u003c/p\u003e"},{"header":"Research problem","content":"\u003cp\u003eWhile interest in Arabic language learning is growing, non-native adult learners\u0026mdash;especially those from diverse cultural and professional backgrounds\u0026mdash;continue to face challenges in mastering Arabic\u0026rsquo;s complex phonology, grammar, and script (AlAfnan, 2025; Alsaied, 2024). Traditional instruction, often delivered through fixed online curricula, lacks the personalization needed to address varied learner needs, particularly in multicultural contexts like that of Greek-speaking learners across Europe and the Gulf.\u003c/p\u003e\n\u003cp\u003eArtificial intelligence (AI) tools\u0026mdash;such as intelligent tutoring systems, catboats, and adaptive learning platforms\u0026mdash;have shown promise in enhancing engagement, adaptively, and language proficiency in second language education (Son et al., 2023; Pokrivč\u0026aacute;kov\u0026aacute;, 2019). However, their application in Arabic learning remains limited and fragmented, often hindered by infrastructural gaps, insufficient teacher training, and ethical concerns (Pokrivč\u0026aacute;kov\u0026aacute;, 2019).\u003c/p\u003e\n\u003cp\u003eThis study addresses the gap by exploring how AI can be effectively used to support Arabic language learning for non-native speakers in diverse, hybrid learning environments.\u003c/p\u003e\n\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n\u003ch2\u003eResearch purpose\u003c/h2\u003e\n\u003cp\u003eThe purpose of this study is to explore how artificial intelligence (AI) can enhance engagement, adaptively, and linguistic proficiency in Arabic language learning among non-native speakers in diverse cultural and geographic contexts. Focusing on a group of Greek-speaking adult learners enrolled in Arabic programs across Saudi Arabia, the UAE, Greece, and other European countries, the study examines how AI tools address individual learning needs, support pronunciation and phonetic accuracy, and personalize instruction across varying proficiency levels (A1\u0026ndash;C2).\u003c/p\u003e\n\u003cp\u003eGiven the challenges associated with Modern Standard Arabic and the learners' wide-ranging motivations\u0026mdash;professional, academic, cultural, and humanitarian\u0026mdash;the study investigates the role of AI in simulating real-life communication, reducing language anxiety, and fostering emotional and cognitive engagement. Ultimately, this research seeks to generate empirical insights into the effective and ethical integration of AI in Arabic language education for non-native learners, particularly within hybrid and virtual learning environments.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eResearch questions\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003c/div\u003e\n\u003cp\u003e1. How do non-native learners describe their experiences with learning Arabic using generative AI tools?\u003c/p\u003e\n\u003cp\u003e2. In what ways do generative AI tools provide personalized and adaptive support for Arabic language learners?\u003c/p\u003e\n\u003cp\u003e3. What challenges do learners face when using generative AI tools for Arabic language learning?\u003c/p\u003e\n\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\n\u003ch2\u003eRelated studies\u003c/h2\u003e\n\u003cp\u003eGamification offers non-native Arabic learners interactive, game-like experiences\u0026mdash;using points, badges, leaderboards, and streaks\u0026mdash;to lower anxiety and boost motivation (Almelhes, \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e; Kotob \u0026amp; Ibrahim, \u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e; Sahrir \u0026amp; Alias, 2011). Smartphone-based games have been shown to enhance speaking skills (Kenali et al., \u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e), and touchscreen apps improve letter-form recognition (Al Hejaili \u0026amp; Newbury, 2023). Culturally relevant game content further strengthens comprehension and retention (Saleh, Arifin, \u0026amp; Hanefarezan, 2022). Platforms like Duolingo deliver immediate rewards that reinforce foundational vocabulary and grammar, though they may struggle with advanced communicative tasks (Loewen et al., \u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e\n\u003cp\u003eAdaptive AI systems mark a significant evolution from static gamified approaches by offering real-time personalization that caters to individual learning needs. These systems continuously analyze learners\u0026rsquo; performance and automatically adjust instructional content, providing a tailored learning trajectory that is particularly vital for mastering Arabic\u0026rsquo;s unique challenges in pronunciation, grammar, and script memorization (Putri et al., 2021; Riwanda et al., 2021). Nasaruddin (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e) reported that AI-driven platforms, including those integrating ChatGPT, can generate customized exercises that evolve as learners progress. This adaptive feedback mechanism not only maintains high levels of engagement but also supports self-regulated learning, allowing learners to focus on their specific weaknesses\u0026mdash;a critical improvement over one-size-fits-all gamification models.\u003c/p\u003e\n\u003cp\u003eIntelligent tutoring systems (ITS) and automated writing evaluation tools form the backbone of many AI-driven approaches in language education. ITS platforms simulate one-on-one tutoring by using machine learning algorithms to provide immediate, adaptive feedback on tasks such as sentence construction and vocabulary drills (Putri et al., 2021). Research shows that these systems help learners internalize complex grammatical structures by adjusting task difficulty in real time (Nasaruddin, \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e). Similarly, automated writing evaluation systems use natural language processing (NLP) algorithms to detect errors in syntax, morphology, and style in learners\u0026rsquo; Arabic texts, offering instantaneous corrective feedback that enhances writing fluency and accuracy (Lu, \u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e; Zhao \u0026amp; Zhang, \u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e). Together, these tools reduce the workload on instructors while fostering a self-directed learning environment where learners can repeatedly practice and refine their language skills.\u003c/p\u003e\n\u003cp\u003eConversational agents and chatbots have emerged as a promising means of providing interactive dialogue practice. These AI-driven tools are designed to simulate natural conversation and offer a safe, low-pressure environment where learners can experiment with new vocabulary and sentence structures (Fryer \u0026amp; Carpenter, \u003cspan class=\"CitationRef\"\u003e2006\u003c/span\u003e; Thompson et al., 2018). Nasaruddin (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e) noted that such systems increase learners\u0026rsquo; willingness to engage in spontaneous conversation, while studies by Dokukina and Gumanova (\u003cspan class=\"CitationRef\"\u003e2020\u003c/span\u003e) confirm that chatbots help reduce communication anxiety by eliminating the fear of negative judgment. In the context of Arabic\u0026mdash;a language with significant phonological and syntactic complexities\u0026mdash;such immediate feedback is invaluable for improving fluency and pronunciation. Speech recognition features, now increasingly integrated into these systems, further allow learners to practice oral skills and receive real-time pronunciation corrections (Fathi et al., \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\n\u003cp\u003eChatGPT, a state-of-the-art conversational agent developed using deep learning and advanced NLP techniques, has drawn considerable attention for its potential in language education. ChatGPT is capable of generating contextually appropriate and coherent responses, making it ideal for generating tailored educational materials, interactive practice dialogues, and immediate corrective feedback (Nasaruddin, \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e). Its ability to simulate \u0026ldquo;human-like\u0026rdquo; conversation offers non-native speakers an \u0026ldquo;anxiety-free\u0026rdquo; space to practice and build confidence in their communicative skills (Fryer \u0026amp; Carpenter, \u003cspan class=\"CitationRef\"\u003e2006\u003c/span\u003e; Thompson et al., 2018). Although some studies have identified limitations\u0026mdash;such as occasional generic responses or a lack of depth in sustained interactions\u0026mdash;research suggests that these shortcomings can be mitigated by incorporating structured, teacher-guided activities alongside ChatGPT use (Kasneci et al., \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e). This blended approach leverages ChatGPT\u0026rsquo;s strengths while ensuring that cultural nuances and advanced linguistic subtleties are adequately addressed.\u003c/p\u003e\n\u003cp\u003eThe successful integration of AI technologies into Arabic language instruction depends heavily on the digital competencies and pedagogical readiness of educators. Studies indicate that teachers who are well-prepared and maintain a positive attitude toward technology are more effective in implementing AI-based tools (Al-Bulushi \u0026amp; Al‐Issa, \u003cspan class=\"CitationRef\"\u003e2017\u003c/span\u003e). Comprehensive professional development programs are critical; Pokrivcakova (2019) argues that such training should address both the operational aspects of AI systems and strategies for integrating automated feedback into the broader pedagogical framework. Research by Sto\u0026scaron;ić and Guill\u0026eacute;n-G\u0026aacute;mez (2024) and Nguyen (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e) underscores that teacher training not only improves educators\u0026rsquo; technological literacy but also enhances student engagement and learning outcomes. Blended learning models that combine AI-driven activities with traditional, teacher-led instruction have been shown to yield the most positive results, as they allow educators to contextualize digital interactions with cultural and situational insights.\u003c/p\u003e\n\u003cp\u003eDespite the many benefits of AI-driven language instruction, significant challenges remain. Robust technological infrastructure is essential, yet many educational institutions\u0026mdash;particularly in under-resourced settings\u0026mdash;struggle to afford the advanced hardware and high-speed internet necessary to support these systems (Azmi \u0026amp; Zakaria, \u003cspan class=\"CitationRef\"\u003e2020\u003c/span\u003e; Wulantari et al., \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e). Moreover, AI systems often face inherent limitations in capturing the full nuances of human language. While automated tools effectively correct grammatical errors, they sometimes fall short of providing the contextual, cultural, and stylistic feedback required for proficient Arabic communication (Zhao \u0026amp; Zhang, \u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e; Fryer \u0026amp; Carpenter, \u003cspan class=\"CitationRef\"\u003e2006\u003c/span\u003e).\u003c/p\u003e\n\u003cp\u003eAnother critical challenge is the sensitivity of AI systems to input accuracy. Minor spelling or grammatical errors can trigger inappropriate feedback, potentially frustrating learners and detracting from the overall learning experience (Coniam, \u003cspan class=\"CitationRef\"\u003e2014\u003c/span\u003e). Ethical considerations\u0026mdash;including data privacy, algorithmic bias, and equitable access\u0026mdash;further complicate the integration of AI into language education. Pokrivcakova (2019) cautions that without rigorous ethical safeguards, these technologies may inadvertently reinforce existing inequities or foster culturally insensitive practices. Addressing these issues requires not only technological refinement but also the development of comprehensive policies and infrastructure investments.\u003c/p\u003e\n\u003cp\u003eComparative assessments of IT tools in language education offer valuable insights into the relative strengths and weaknesses of different approaches. Sto\u0026scaron;ić and Guill\u0026eacute;n-G\u0026aacute;mez (2024) conducted a comparative analysis of several IT tools\u0026mdash;including Duolingo, Rosetta Stone, Edmodo, and Tandem\u0026mdash;and found that each tool offers unique advantages that address different facets of language learning. For example, Duolingo\u0026rsquo;s gamified approach excels in boosting engagement and facilitating vocabulary acquisition for beginners (Loewen et al., \u003cspan class=\"CitationRef\"\u003e2019\u003c/span\u003e), whereas Rosetta Stone\u0026rsquo;s immersive framework is particularly effective in developing listening and pronunciation skills (Lord, \u003cspan class=\"CitationRef\"\u003e2015\u003c/span\u003e). Collaborative platforms like Edmodo foster group interaction and critical thinking, and Tandem, which connects learners with native speakers, has been shown to improve speaking fluency and intercultural competence. These findings suggest that no single tool can address all aspects of language learning; rather, a multimodal, hybrid approach is necessary to meet the diverse needs of non-native Arabic speakers (Dokukina \u0026amp; Gumanova, \u003cspan class=\"CitationRef\"\u003e2020\u003c/span\u003e; Woo \u0026amp; Choi, \u003cspan class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e\n\u003cp\u003eEmerging technologies such as augmented reality (AR) and virtual reality (VR) offer exciting new possibilities for Arabic language instruction by creating immersive learning environments that simulate authentic cultural contexts. Valencia et al. (\u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e) and Panagiotidis (\u003cspan class=\"CitationRef\"\u003e2021\u003c/span\u003e) have shown that AR/VR applications can enhance both linguistic competence and intercultural understanding by providing realistic simulations of everyday interactions. Although research specific to Arabic is still in its early stages, promising findings from other language contexts (Wang et al., 2019; Viktoria et al., 2018) suggest that these immersive technologies can be effectively adapted to teach Arabic.\u003c/p\u003e\n\u003cp\u003eMobile learning applications further extend the accessibility and flexibility of language education. Gafni et al. (\u003cspan class=\"CitationRef\"\u003e2017\u003c/span\u003e) and Talan et al. (\u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e) have documented that mobile platforms\u0026mdash;by combining gamification with AI-driven adaptive learning\u0026mdash;enable learners to access personalized content anytime and anywhere. These applications break down geographical and temporal barriers, allowing non-native Arabic speakers to engage with learning materials on the go. Moreover, mobile learning fosters continuous, self-directed study, a feature that is particularly beneficial in modern educational contexts where flexibility and personalization are key.\u003c/p\u003e\n\u003cp\u003eA holistic, multimodal approach that integrates various IT tools and emerging technologies is essential for addressing the complex challenges of teaching Arabic to non-native speakers. Evidence from the literature suggests that gamified platforms are most effective for early-stage vocabulary acquisition and engagement, while immersive systems like Rosetta Stone bolster receptive skills through context-driven practice. Collaborative tools such as Edmodo enhance communicative competence by fostering group interaction, and real-world conversation platforms like Tandem significantly improve speaking fluency and intercultural competence (Sto\u0026scaron;ić \u0026amp; Guill\u0026eacute;n-G\u0026aacute;mez, 2024; Dokukina \u0026amp; Gumanova, \u003cspan class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e\n\u003cp\u003eFurthermore, the integration of AR and VR could provide the cultural context and immersive experiences necessary for comprehensive language acquisition, complementing the adaptive strengths of AI-driven systems (Panagiotidis, \u003cspan class=\"CitationRef\"\u003e2021\u003c/span\u003e; Valencia et al., \u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e). To fully harness these benefits, continuous teacher professional development is imperative. Educators must receive ongoing training to stay current with technological advancements and to integrate these tools effectively into a blended instructional model (Al-Bulushi \u0026amp; Al‐Issa, \u003cspan class=\"CitationRef\"\u003e2017\u003c/span\u003e; Pokrivcakova, 2019). Finally, addressing infrastructural and ethical challenges\u0026mdash;by investing in robust technology, developing rigorous privacy policies, and ensuring equitable access\u0026mdash;is essential for the sustainable integration of AI in Arabic language education.\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003eContext of the study\u003c/h3\u003e\n\u003cp\u003eA comprehensive study of AI in Arabic education involves 19\u0026ndash;59-year-old Greek learners spread across Saudi Arabia, the UAE, Greece, and other European countries. Their proficiency ranges from A1 to C2, creating a culturally and linguistically diverse learning environment. Instruction follows a fixed textbook aligned with the curriculum and is delivered via Zoom: the syllabus appears onscreen, discussions occur entirely in Arabic, and interactive exercises simulate real-world scenarios.\u003c/p\u003e\n\u003cp\u003eLearners\u0026rsquo; motivations vary: some seek career advancement\u0026mdash;especially in Gulf job markets\u0026mdash;while others pursue cultural understanding, academic interests in Middle Eastern studies or comparative religion, or personal reasons such as family ties. In Greece, many students are police, military, or human-rights workers aiming to communicate with migrants from Syria, Palestine, Iraq, and Libya; despite diverse dialects, they often focus on Modern Standard Arabic as a common linguistic medium.\u003c/p\u003e\n\u003cp\u003eAI tools create virtual environments that bridge abstract learning and lived experience, fostering emotional engagement with language through realistic simulations. They also offer accurate pronunciation models and feedback on subtle phonetic contrasts\u0026mdash;soft versus emphatic, voiced versus voiceless, plosive versus non-plosive\u0026mdash;that non-native speakers frequently confuse. By combining these technological affordances with interactive, instructor-led activities, AI enhances both the cognitive and affective dimensions of Arabic acquisition.\u003c/p\u003e"},{"header":"Methodology","content":"\u003cp\u003eWe adopted a qualitative research approach to investigate how generative AI tools are being used to support Arabic language learning among non-native speakers. A qualitative approach is particularly appropriate for exploring how individuals experience and interpret emerging educational technologies within their specific learning contexts (Mohajan, 2018). To guide our inquiry, we employed a case study design, which allows for an in-depth examination of a real-life phenomenon within its natural setting. This design enabled us to explore the integration of generative AI in Arabic language instruction from the perspective of diverse learners across cultural and geographical contexts (Yin, 2003).\u003c/p\u003e\u003ch3\u003eParticipant\u003c/h3\u003e\u003cp\u003eThe study involved 25 Greek-speaking adult learners of Arabic, ranging in age from 19 to 59 years. This diverse cohort was geographically dispersed across Saudi Arabia, the United Arab Emirates, Greece, and other European countries, reflecting a rich blend of cultural and professional backgrounds. Participants exhibited varying levels of Arabic language proficiency, from beginner (A1) to advanced (C2), creating an inclusive learning environment that accommodated a wide range of linguistic abilities. Instruction was delivered via Zoom using a standardized textbook aligned with the formal Arabic language curriculum for non-native speakers. Classes were conducted entirely in Arabic and emphasized interactive, real-world application through exercises and discussions.\u003c/p\u003e\u003cp\u003eParticipants’ motivations for learning Arabic were multifaceted. Some aimed to enhance career opportunities, particularly in Gulf countries, while others pursued the language for academic, cultural, or personal reasons—such as deepening their understanding of Arab societies, supporting human rights work, or connecting with Arab family members or partners. In Greece specifically, many learners were military personnel or police officers seeking to communicate more effectively with migrants from Arabic-speaking regions.\u003c/p\u003e\u003ch3\u003eRecruiting the participants\u003c/h3\u003e\u003cp\u003eParticipants in this study were selected based on specific criteria aligned with the research objectives. All participants were non-native speakers of Arabic, with Greek as their first language, and were either currently enrolled in or had recently completed formal Arabic language courses designed for non-native speakers. They ranged in age from 19 to 59 years, ensuring a broad representation of adult learners. To capture diverse experiences with Arabic language education, participants were drawn from various geographic locations, including Saudi Arabia, the United Arab Emirates, Greece, and other European countries. Their proficiency in Arabic varied widely—from beginner (A1) to advanced (C2)—which allowed the study to examine how generative AI tools support learners at different levels. The selection also considered the participants’ motivations for studying Arabic, which ranged from professional development and academic interest to humanitarian work and personal or cultural reasons. Although prior use of generative AI tools such as ChatGPT, Google Translate, or Duolingo was not a strict requirement, participants with relevant experience were prioritized to ensure meaningful insights into the role of AI in Arabic language learning.\u003c/p\u003e\u003ch2\u003eData collection\u003c/h2\u003e\u003cp\u003eTo collect qualitative data for this study, we used an open-ended questionnaire designed to capture participants’ lived experiences with using generative AI tools in Arabic language learning. This method allowed for in-depth exploration of participants' perceptions, practices, and challenges, as they articulated their responses in their own words without constraints. Open-ended questions are widely recognized for their ability to elicit rich, detailed insights into individuals’ experiences and attitudes (Creswell, 2014; Harlacher, 2016). According to Thorndike and Thorndike-Christ (2010), open-ended items empower participants to express their thoughts freely, enabling researchers to gather nuanced information on complex phenomena.\u003c/p\u003e\u003cp\u003eThe questions used in the form were developed based on findings from prior research on technology use in education, particularly the works of Joo et al. (2016), Çoklar et al. (2017), Özgür (2020), and Christian et al. (2020), and were adapted to the context of generative AI and language learning. The final version of the form was written in Arabic and distributed to a diverse group of Greek-speaking adult learners of Arabic, all of whom were over the age of 18.\u003c/p\u003e\u003cp\u003eThe questionnaire consisted of three sections. The first section provided participants with an overview of the study and outlined their rights as voluntary participants. The second section collected demographic information. The third and main section contained ten open-ended questions focusing on the participants' experiences with generative AI tools—such as ChatGPT, Duolingo, and Google Translate—and how these tools have influenced their Arabic language learning journey. The questions addressed key areas such as AI-assisted vocabulary development, grammar support, pronunciation practice, user engagement, perceived challenges, and comparisons with traditional instructional methods.\u003c/p\u003e\u003cp\u003eThe form was designed using Microsoft Office 365 and distributed via email. This platform was selected because of its accessibility and user familiarity; all participants had access to Office 365 and were accustomed to using its features, ensuring ease of response submission.\u003c/p\u003e\u003ch2\u003eData analysis\u003c/h2\u003e\u003cp\u003eWe conducted a thematic analysis of open-ended questionnaire data from non-native Arabic learners using generative AI tools, allowing themes to emerge organically from their experiences (Creswell, 2014; Drisko \u0026amp; Maschi, 2016). Using NVivo 14, we followed Hsieh and Shannon’s (2005) three-phase procedure. First, we prepared the data—numbering each response, consolidating into a single file, and cleaning for coding. Second, we identified units of analysis (sentences or phrases reflecting learners’ perceptions), informed by AI-in-education research (Joo et al., 2016; Özgür, 2020), and developed an initial codebook from the literature and recurrent data patterns. A 10% pre-test by an external researcher refined the codes. In the third phase, two researchers independently applied the finalized codebook, achieving 86% intercoder agreement; disagreements were resolved through consensus (Creswell \u0026amp; Miller, 2000). This systematic, collaborative approach ensured credible identification of themes related to engagement, adaptivity, linguistic challenges, and perceived AI effectiveness.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eResearch Question #1\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eHow do non-native learners describe their experiences with learning Arabic using generative AI tools?\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe aim of the first research question was to explore learners\u0026rsquo; experiences with studying Arabic as a non-native language and to understand the challenges they encountered throughout their learning journey. The 25 participants in this study\u0026mdash;Greek-speaking adult learners residing across Europe and the Gulf region\u0026mdash;shared diverse perspectives shaped by their linguistic backgrounds, learning environments, and personal motivations. Their responses revealed a rich blend of enthusiasm, cultural appreciation, and perseverance, but also highlighted persistent challenges such as mastering the Arabic script, adapting to unfamiliar grammar structures, and accessing high-quality instructional support. From their narratives, five major themes emerged: positive learning experiences, motivational factors, learning challenges, learning environments, and the early stages of acquisition. Each theme included several subthemes reflecting learners\u0026rsquo; personal goals, emotional engagement, and the contextual barriers they face. Table \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e presents a synthesis of these themes, subthemes, and illustrative quotations drawn from participants\u0026rsquo; lived experiences.\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n \u003cdiv class=\"colspec\" align=\"left\"\u003e\u0026nbsp;\u003c/div\u003e\n \u003ctable id=\"Tab1\" border=\"1\"\u003e\n \u003ccaption\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eThemes and Subthemes for RQ1: Learners\u0026rsquo; Experience with Learning Arabic\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eMain Theme\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eSubtheme\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eSupporting Quotations\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePositive Learning Experience\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eEnjoyment in Learning Arabic\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eParticipants highlighted that \u0026quot;My experience so far has been great, I enjoy learning languages.\u0026quot; (S1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCultural Appreciation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eParticipants highlighted that \u0026quot; The Arabic language is interesting because of its symbols and its differences from other languages.\u0026quot; (S2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMotivational Factors\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eProfessional Needs\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eParticipants highlighted that \u0026quot;I am working in Saudi Arabia, so I wanted to do this step for professional and personal reasons.\u0026quot; (S12)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePersonal Interest\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSeveral learners emphasized \u0026quot;I am working in Saudi Arabia, so I wanted to do this step for professional and personal reasons.\u0026quot; (S7)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eChallenges in Learning\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAlphabet and Script\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLearners frequently mentioned \u0026quot; \u0026quot;The learning pace was time-consuming, and the instruction was not oriented toward obtaining any certification.\u0026quot; (S3)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLack of Certified Instruction\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSome responses revealed \u0026quot;It is a language that is unique, so you have to learn everything from zero.\u0026quot; S6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLearning Environment\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSmall Group Learning\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIt was noted that \u0026quot;I am in a small group teaching course.\u0026quot; (Participant s4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIndependent Study\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIt was noted that \u0026quot;I am in a small group teaching course.\u0026quot; (S7)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eEarly Learning Stage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLimited Experience\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIt was noted that \u0026quot;I\u0026rsquo;m just beginning my journey (two sessions in), so I don\u0026rsquo;t have all the full picture yet.\u0026quot; (S8)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eInitial Enthusiasm\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eA number of participants shared that \u0026quot;I\u0026rsquo;m just beginning my journey (two sessions in), so I don\u0026rsquo;t have all the full picture yet.\u0026quot; (S12)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\n \u003ch2\u003ePositive Learning Experience\u003c/h2\u003e\n \u003cp\u003eSome participants described their experience learning Arabic as highly enjoyable and personally fulfilling. For example, one learner shared, \u003cem\u003e\u0026quot;My experience so far has been great, I enjoy learning languages.\u0026quot;\u003c/em\u003e (S1). Others expressed a deep appreciation for the language\u0026rsquo;s cultural uniqueness, such as Participant S3 who noted, \u003cem\u003e\u0026quot;The Arabic language is interesting because of its symbols and its differences from other languages.\u0026quot;\u003c/em\u003e\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\n \u003ch2\u003eMotivational Factors\u003c/h2\u003e\n \u003cp\u003eMotivations for learning Arabic varied, but several participants emphasized both personal and professional goals. One participant, for instance, explained, \u003cem\u003e\u0026quot;I am working in Saudi Arabia, so I wanted to do this step for professional and personal reasons.\u0026quot;\u003c/em\u003e (S7). This reflects how learners are driven not only by career advancement but also by a genuine interest in engaging with the local culture and language.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\n \u003ch2\u003eChallenges in Learning\u003c/h2\u003e\n \u003cp\u003eDespite their enthusiasm, learners reported notable challenges. A few participants found the Arabic alphabet and script especially difficult to master, with one commenting, \u003cem\u003e\u0026quot;The Arabic language is interesting because of its symbols and its differences from other languages.\u0026quot;\u003c/em\u003e (S2). Others noted the lack of structured or certified instruction as a barrier, as one participant described, \u003cem\u003e\u0026quot;It is a language that is unique, so you have to learn everything from zero.\u0026quot;\u003c/em\u003e (S9)\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\n \u003ch2\u003eLearning Environment\u003c/h2\u003e\n \u003cp\u003eParticipants described varying learning environments, with some emphasizing the benefits of small group instruction. As one noted, \u003cem\u003e\u0026quot;I am in a small group teaching course.\u0026quot;\u003c/em\u003e (S4). Others reported a more self-directed learning experience, as reflected in another comment: \u003cem\u003e\u0026quot;I am in a small group teaching course.\u0026quot;\u003c/em\u003e (S11). While the contexts differed, these environments shaped how learners interacted with the language and the AI tools.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\n \u003ch2\u003eEarly Learning Stage\u003c/h2\u003e\n \u003cp\u003eSeveral participants were at the beginning of their Arabic learning journey and acknowledged their limited exposure so far. One learner remarked, \u003cem\u003e\u0026quot;I\u0026rsquo;m just beginning my journey (two sessions in), so I don\u0026rsquo;t have all the full picture yet.\u0026quot;\u003c/em\u003e (S5). This sense of initial enthusiasm combined with uncertainty was common among early-stage learners as they navigated the complexities of Arabic for the first time.\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003eResearch Question 2\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cem\u003eIn what ways do generative AI tools provide personalized and adaptive support for Arabic language learners?\u003c/em\u003e\u003c/p\u003e\n \u003cp\u003eThe second research question aimed to examine how generative AI tools support personalized and adaptive learning experiences for non-native Arabic learners across varying proficiency levels. The participants described how tools such as Google Translate, ChatGPT, and Duolingo provided flexible, real-time assistance that catered to their individual learning needs. Their responses revealed six main themes related to adaptivity: personalized vocabulary support, pronunciation and listening features, self-paced learning and flexibility, grammar and sentence structure, learning confidence and control, and limitations in personalization. Within these themes, learners emphasized the usefulness of AI features such as synonym suggestions, audio pronunciation playback, and the ability to study at their own pace and on their own schedule. However, participants also noted shortcomings, particularly in handling Arabic dialects, delivering nuanced grammatical support, and providing culturally sensitive feedback. Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e summarizes the themes, subthemes, and participant quotations that reflect the multifaceted role of AI in delivering adaptive and learner-centered support in Arabic language acquisition.\u003c/p\u003e\n \u003cdiv class=\"gridtable\"\u003e\n \u003cdiv class=\"colspec\" align=\"left\"\u003e\u0026nbsp;\u003c/div\u003e\n \u003cdiv class=\"colspec\" align=\"left\"\u003e\u0026nbsp;\u003c/div\u003e\n \u003cdiv class=\"colspec\" align=\"left\"\u003e\u0026nbsp;\u003c/div\u003e\n \u003ctable id=\"Tab2\" border=\"1\"\u003e\n \u003ccaption\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eThemes and Subthemes for RQ2: Adaptively and Personalized Support through AI Tools\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eMain Theme\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eSubtheme\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eSupporting Quotations\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePersonalized Vocabulary Support\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eWord Suggestions\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLearners frequently mentioned \u0026quot;Google Translate proposes other words that have the same meaning, which helps deepen my Arabic knowledge.\u0026quot; (S1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSynonym Alternatives\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSeveral learners emphasized \u0026quot; \u0026quot;It would form a sample sentence in English (S12)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eVisual Reinforcement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSeveral learners emphasized \u0026quot;Duolingo helps a lot, especially for vocabulary.\u0026quot; (S25)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePronunciation and Listening Features\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAudio Playback\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIt was noted that \u0026quot;I have found AI really helpful regarding pronunciation.\u0026quot; (Participant s1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePhonetic Transcription\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIt was noted that \u0026quot; \u0026quot;It cannot recognize the dialect and suggests words that are far from the appropriate answer.\u0026quot; (S17)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRepetition Support\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSeveral learners emphasized \u0026quot;I use it to hear how the word sounds.\u0026quot; (S21)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSelf-Paced Learning and Flexibility\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIndependent Learning\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eParticipants highlighted that \u0026quot;There are gaps in support for the Greek language.\u0026quot; (S19)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLearning Anytime\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eParticipants highlighted that \u0026quot;AI allows fast access to information and boosts productivity.\u0026quot; (S23)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePacing According to Need\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSeveral learners emphasized \u0026quot;I think if you use it for specific words it can be helpful to move on in an exercise.\u0026quot; (S22)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGrammar and Sentence Structure\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eRoot Word Discovery\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eParticipants highlighted that \u0026quot;I usually use it to find the root of the verbs.\u0026quot; (S22)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGrammar Assistance\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eA number of participants shared that \u0026quot;I used it to form a sentence as an example in the English language. (S16))\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSentence Examples\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSeveral learners emphasized \u0026quot;ChatGPT is used mostly for document translation in my job.\u0026quot; (S5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLimitations in Personalization\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDialect Recognition Issues\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eA number of participants shared that \u0026ldquo;one of its limitations is It cannot recognize the dialect and suggests words that deviate from the accurate answer (S11)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLow Support for Greek\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSome responses revealed \u0026quot;There are deficiencies in the Greek language support\u0026quot; (S14)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGeneralized Feedback\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eParticipants highlighted that \u0026quot;Sometimes when translating a sentence, the meaning makes no sense.\u0026quot; (S3)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLearning Confidence and Control\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAssurance When Stuck\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLearners frequently mentioned \u0026quot;It doesn\u0026rsquo;t make me feel more confident, but it makes me feel secure if I get stuck.\u0026quot; (S10)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eReduced Fear of Mistakes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eParticipants highlighted that \u003cstrong\u003e\u0026quot;\u003c/strong\u003eFor me, there is no fear of failure of using it [AI tools] (S6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eConfidence Building Over Time\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eIt was noted that \u0026quot;Generative AI can be helpful as a tool in general for all those fields.\u0026quot; (S12)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e\n \u003ch2\u003ePersonalized Vocabulary Support\u003c/h2\u003e\n \u003cp\u003eMany participants appreciated the way AI tools enhanced their vocabulary development. For example, one learner explained, \u0026quot;Google Translate proposes other words that have the same meaning, which helps deepen my Arabic knowledge.\u0026quot; (S1). Another participant noted how AI tools provided semantic alternatives and explanations, stating, \u0026quot;It often presented words with related meanings and explained the meaning of the word in a descriptive way.\u0026quot; (S5). In addition, several learners pointed to the effectiveness of Duolingo in reinforcing vocabulary through repetition and visuals, with one sharing, \u0026quot;Duolingo helps a lot, especially for vocabulary.\u0026quot; (S11)\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec20\" class=\"Section2\"\u003e\n \u003ch2\u003ePronunciation and Listening Features\u003c/h2\u003e\n \u003cp\u003eParticipants emphasized the usefulness of AI in supporting pronunciation and listening comprehension. One learner shared, \u0026quot;I have found AI really helpful regarding pronunciation.\u0026quot; (S14), while another highlighted the benefit of auditory repetition: \u0026quot;These tools also have the option for the learner to hear the word.\u0026quot; (S18). Similarly, one participant explained, \u0026quot;I use it to hear how the word sounds.\u0026quot; (S17), indicating how repetition can enhance phonetic awareness.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec21\" class=\"Section2\"\u003e\n \u003ch2\u003eSelf-Paced Learning and Flexibility\u003c/h2\u003e\n \u003cp\u003eSeveral participants emphasized the flexibility AI tools offer for independent and self-directed learning. For instance, one noted, \u0026quot;It [AI tools] promotes equal opportunities and the learner\u0026rsquo;s self-initiative while learning new languages.\u0026quot; (S21). Others appreciated the immediate access to information, as expressed by Participant S25: \u0026quot;AI allows fast access to information and boosts productivity.\u0026quot; Additionally, some learners reported that AI helped them progress at their own pace, with one stating, \u0026quot;I think if you use it for specific words it can be helpful to move on in an exercise.\u0026quot; (S13)\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec22\" class=\"Section2\"\u003e\n \u003ch2\u003eGrammar and Sentence Structure\u003c/h2\u003e\n \u003cp\u003eParticipants also described how AI tools supported their understanding of Arabic grammar and sentence construction. One learner reflected, \u0026quot;It helped me try to identify the root forms of verbs, which supported my understanding of how the language is structured.\u0026quot; (S23). Another noted that AI tools provided examples in English that aided comprehension, saying, \u0026quot;It would form a sample sentence in English.\u0026quot; (S12). In workplace contexts, some participants reported using AI primarily for functional translation tasks, as one stated, \u0026quot;ChatGPT is used mostly for document translation in my job.\u0026quot; (S15)\u003c/p\u003e\n \u003cdiv id=\"Sec23\" class=\"Section3\"\u003e\n \u003ch2\u003eLimitations in Personalization\u003c/h2\u003e\n \u003cp\u003eDespite the benefits, participants also highlighted key limitations in AI personalization. Several learners noted that dialect recognition remains a major challenge, with one stating, \u0026quot;It cannot recognize the dialect and suggests words that are far from the appropriate answer.\u0026quot; (S17). Others pointed to insufficient language support, particularly in Greek, explaining, \u0026quot;There are gaps in support for the Greek language.\u0026quot; (S19). A few participants also expressed frustration with generalized or inaccurate feedback, such as: \u0026quot;Sometimes when translating a sentence, the meaning makes no sense.\u0026quot; (S21)\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec24\" class=\"Section2\"\u003e\n \u003ch2\u003eLearning Confidence and Control\u003c/h2\u003e\n \u003cp\u003eParticipants reported mixed experiences regarding how AI tools influenced their confidence in language use. One participant expressed a sense of security rather than confidence, stating, \u0026quot;It doesn\u0026rsquo;t make me feel more confident, but it makes me feel secure if I get stuck.\u0026quot; (S6). Others found reassurance in the learning process, such as one who noted, \u0026quot;There is no fear of failure.\u0026quot; (S3). Additionally, some participants recognized AI\u0026rsquo;s potential as a supportive tool over time, with one explaining, \u0026quot;Generative AI can be helpful as a tool in general for all those fields.\u0026quot; (S4)\u003c/p\u003e\n \u003cp\u003eResearch Question #3: \u003cem\u003eWhat challenges do learners face when using generative AI tools for Arabic language learning?\u003c/em\u003e\u003c/p\u003e\n \u003cp\u003eThe third research question focused on identifying the challenges and limitations that non-native learners encounter when using generative AI tools to support their Arabic language learning. Participants reported a variety of concerns, which we organized into five key themes: translation inaccuracy, dialect and cultural nuance limitations, dependence and reduced critical thinking, limited customization, and language support inequities. Learners frequently noted that AI-generated translations often failed to capture the intended meaning, especially when translating full sentences or culturally rich expressions. Others expressed frustration with the tools\u0026rsquo; inability to differentiate between dialects or provide accurate context-based feedback. Additionally, several participants reported a growing dependence on AI for tasks such as spelling or grammar correction, which led to reduced memory recall and critical language processing. Concerns were also raised about the limited support for the Greek language and the dominance of English in AI responses. Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e presents these themes and subthemes along with selected participant quotations that illustrate the barriers encountered in using AI for Arabic language learning.\u003c/p\u003e\n \u003cdiv class=\"gridtable\"\u003e\n \u003cdiv class=\"colspec\" align=\"left\"\u003e\u0026nbsp;\u003c/div\u003e\n \u003cdiv class=\"colspec\" align=\"left\"\u003e\u0026nbsp;\u003c/div\u003e\n \u003cdiv class=\"colspec\" align=\"left\"\u003e\u0026nbsp;\u003c/div\u003e\n \u003ctable id=\"Tab3\" border=\"1\"\u003e\n \u003ccaption\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eThemes and Subthemes for RQ3: Challenges and Limitations of Using AI Tools in Arabic Learning\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr style=\"height: 35px;\"\u003e\n \u003cth style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eMain Theme\u003c/p\u003e\n \u003c/th\u003e\n \u003cth style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eSubtheme\u003c/p\u003e\n \u003c/th\u003e\n \u003cth style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eSupporting Quotations\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr style=\"height: 35px;\"\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eTranslation Inaccuracy\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eMeaning Distortion\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eIt was noted that \u0026quot;Most of the times I can\u0026rsquo;t sufficiently pass the message that I intend.\u0026quot; (S7)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr style=\"height: 35px;\"\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eGrammar Errors\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eIt was noted that \u0026quot;Sometimes when you try to translate a sentence, the meaning will make no sense.\u0026quot; (S13)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr style=\"height: 48px;\"\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\n \u003cp\u003eLack of Context\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\n \u003cp\u003eSeveral learners emphasized \u0026quot;I\u0026rsquo;ve noticed that translations into Greek are not always accurate, which can make it harder for me to fully understand the intended meaning.\u0026quot; (S12)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr style=\"height: 48px;\"\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\n \u003cp\u003eDialect and Cultural Nuance Limitations\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\n \u003cp\u003eDialect Misinterpretation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\n \u003cp\u003eLearners frequently mentioned \u0026quot;It struggles to translate proverbs, idiomatic expressions, and dialects accurately, which affects my ability to fully understand cultural and contextual meanings.\u0026quot; (S18)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr style=\"height: 35px;\"\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eIdiom Misunderstanding\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eSeveral learners emphasized \u0026quot;Minor issues exist that an expert translator should finalize.\u0026quot; (S9)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr style=\"height: 35px;\"\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eCultural Gaps\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eIt was noted that \u0026quot;You can\u0026rsquo;t trust it 100%.\u0026quot; (S9)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr style=\"height: 48px;\"\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\n \u003cp\u003eDependence and Reduced Critical Thinking\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\n \u003cp\u003eOverreliance on AI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\n \u003cp\u003eIt was noted that \u003cstrong\u003e\u0026quot;\u003c/strong\u003eI often rely on it to check the spelling of words, which makes me less likely to retain them independently.\u0026quot; \u003cstrong\u003e(S23)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr style=\"height: 48px;\"\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\n \u003cp\u003eMemory Inhibition\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\n \u003cp\u003eIt was noted I don\u0026rsquo;t actively engage my thinking and memory, which leads to noticeable gaps in my learning over time. (S17)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr style=\"height: 35px;\"\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eCopy-Paste Learning\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eLearners frequently mentioned \u0026quot;It\u0026rsquo;s helpful, but it doesn\u0026rsquo;t build confidence.\u0026quot; (S6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr style=\"height: 48px;\"\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\n \u003cp\u003eLimited Customization\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\n \u003cp\u003eLack of Personalized Feedback\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\n \u003cp\u003eParticipants highlighted that \u0026quot;I would like to see more tools focused in specific fields like Geography, science etc.\u0026quot; (S4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr style=\"height: 48px;\"\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\n \u003cp\u003eInsufficient Task Variety\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 48px;\" align=\"left\"\u003e\n \u003cp\u003eSome responses revealed \u0026quot; \u0026quot;It would be beneficial if learners had the option to create their own quizzes and receive clear explanations for any mistakes.(S3)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr style=\"height: 35px;\"\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eTopic Irrelevance\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eSome responses revealed \u0026quot;\u0026quot;it could support classroom learning by providing flashcards with images.\u0026quot; (S10)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr style=\"height: 35px;\"\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eLanguage Support Inequities\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eStronger English Support\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eSome responses revealed \u0026quot;\u003cem\u003ethere seem to be more features available in English.\u0026quot;\u003c/em\u003e (S15)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr style=\"height: 35px;\"\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eWeaker Greek Integration\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eParticipants highlighted \u0026quot;I don\u0026rsquo;t really trust the information given by AI.\u0026quot; (S2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr style=\"height: 35px;\"\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eTool Limitations\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n \u003cp\u003eSome responses revealed \u0026quot;I don\u0026rsquo;t really trust the information given by AI.\u0026quot; (Participant s1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Sec25\" class=\"Section3\"\u003e\n \u003ch2\u003eTranslation Inaccuracy\u003c/h2\u003e\n \u003cp\u003eA number of participants expressed concerns about the limitations of AI-generated translations in conveying accurate meaning. For instance, one learner remarked, \u0026quot;Most of the times I can\u0026rsquo;t sufficiently pass the message that I intend.\u0026quot; (S1). Several participants also emphasized issues related to grammatical accuracy, with one stating, \u0026quot;Sometimes when you try to translate a sentence, the meaning will make no sense.\u0026quot; (S6). Additionally, some participants noted that poor contextual understanding often hindered comprehension, as reflected in the comment: \u0026quot;The translation into Greek is not always accurate, which sometimes affects my understanding of the intended meaning.\u0026quot; (S8)\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Sec26\" class=\"Section3\"\u003e\n \u003ch2\u003eDialect and Cultural Nuance Limitations\u003c/h2\u003e\n \u003cp\u003eSome participants highlighted the inability of AI tools to handle the complexity of dialects and culturally embedded expressions. One participant observed, \u0026quot;It is unable to accurately translate proverbs, idiomatic expressions, and dialects, which limits my ability to grasp the deeper meaning of certain phrases.\u0026quot; (S9). Others noted that outputs often lacked cultural refinement, with one learner stating, \u0026quot;Minor issues exist that an expert translator should finalize.\u0026quot; (S4). A few participants also expressed general distrust, emphasizing that \u0026quot;You can\u0026rsquo;t trust it 100%.\u0026quot; (S15)\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Sec27\" class=\"Section3\"\u003e\n \u003ch2\u003eDependence and Reduced Critical Thinking\u003c/h2\u003e\n \u003cp\u003eSeveral participants raised concerns about becoming overly reliant on AI tools for routine language tasks. One learner admitted, \u0026quot;I often rely on AI to check the spelling of words, which makes me less likely to recall them on my own.\u0026quot; (S18). Others reflected on how this dependence affects cognitive engagement, such as Participant Sa4, who noted, \u0026quot;I don\u0026rsquo;t actively engage my thinking and memory skills, which sometimes results in gaps in my language use.\u0026quot; Additionally, some learners questioned the long-term benefits of AI-supported learning, with one stating, \u0026quot;It\u0026rsquo;s helpful, but it doesn\u0026rsquo;t build confidence.\u0026quot; (S5).\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec28\" class=\"Section2\"\u003e\n \u003ch2\u003eLimited customization\u003c/h2\u003e\n \u003cp\u003eThe majority of participants emphasized the need for domain-specific resources that align with learners\u0026rsquo; academic or professional interests. For example, one participant noted, \u003cem\u003e\u0026quot;I would like to see more tools focused in specific fields like Geography, science, etc.\u0026quot;\u003c/em\u003e (S4). Moreover, a few participants highlighted the importance of interactive features that promote learner autonomy, expressing a desire for \u003cem\u003e\u0026quot;the option to create their own quizzes and receive explanations for their mistakes\u0026quot;\u003c/em\u003e (S16). On the other hand, some participants pointed to the value of visual aids in supporting vocabulary development, suggesting that \u003cem\u003e\u0026quot;it could support classroom learning by providing visual flashcards to help with vocabulary acquisition.\u0026quot;\u003c/em\u003e (S19).\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec29\" class=\"Section2\"\u003e\n \u003ch2\u003eLanguage support inequities\u003c/h2\u003e\n \u003cp\u003eA number of participants observed that generative AI tools tend to offer more robust functionality in English compared to other languages. As one participant noted, \u003cem\u003e\u0026quot;I notice that there are more features available in English.\u0026quot;\u003c/em\u003e (S11). In addition, several participants pointed out limited integration and support for the Greek language, with one learner stating, \u003cem\u003e\u0026quot;There are gaps in support for the Greek language.\u0026quot;\u003c/em\u003e (S7). Some participants also expressed concerns about the reliability of AI-generated content, as reflected in the comment: \u003cem\u003e\u0026quot;I don\u0026rsquo;t really trust the information given by AI.\u0026quot;\u003c/em\u003e (S2).\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eLearners in this study consistently reported that generative AI tools transformed their Arabic studies from a series of isolated drills into an engaging, culturally rich experience. Where traditional gamified apps rely on points or badges to motivate, AI-powered conversational agents simulated realistic dialogues that reduced anxiety and encouraged experimentation with new vocabulary and structures (Almelhes, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Kotob \u0026amp; Ibrahim, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). These interactions often felt more meaningful than rote practice, since learners could immediately see how words and phrases functioned in context. In doing so, AI did not simply supplement instruction; it created its own micro-environments in which emotional engagement with the language flourished, echoing early findings on the low‐stakes affordances of chatbots (Fryer \u0026amp; Carpenter, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2006\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe adaptive feedback mechanisms built into these systems played a crucial role in scaffolding learner progress. Unlike one-size‐fits‐all exercises, AI continuously analyzed individual performance\u0026mdash;suggesting synonyms, adjusting pronunciation models, and calibrating task difficulty to each user\u0026rsquo;s proficiency (Putri et al., 2021; Nasaruddin, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This immediacy enabled learners to target specific weaknesses and to consolidate gains through spaced, iterative practice that mirrored the personalized pathways of intelligent tutoring systems (Putri et al., 2021). In effect, AI became a virtual tutor capable of tailoring instruction in real time, fostering self‐regulated study habits that many learners found both efficient and empowering.\u003c/p\u003e \u003cp\u003eDespite these advances, significant gaps emerged around translation accuracy and cultural nuance. Participants frequently encountered misrendered idioms or literal translations that failed to convey intended meaning\u0026mdash;limitations long documented in natural language processing research (Coniam, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Zhao \u0026amp; Zhang, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). AI\u0026rsquo;s difficulty with regional dialects further underscored the challenge of modeling Arabic\u0026rsquo;s sociolinguistic diversity (Dokukina \u0026amp; Gumanova, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). When tools produced outputs that felt \u0026ldquo;off\u0026rdquo; or culturally insensitive, learners had to rely on external resources or instructor guidance to resolve misunderstandings. These persistent shortcomings highlight the need for richer, more dialect-aware corpora and for collaborative refinement between AI developers and language experts.\u003c/p\u003e \u003cp\u003eA subtler concern centered on cognitive overreliance. While many learners appreciated that AI provided a safety net for spelling, grammar, or pronunciation, several noted that habitual use sometimes weakened their own recall and analytical engagement. This phenomenon has been identified in reviews of chatbot-supported learning, which caution that easy access to answers can inadvertently diminish active problem‐solving and memory consolidation (Huang, Hew, \u0026amp; Fryer, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Without deliberate instructional design to counterbalance AI assistance\u0026mdash;such as tasks requiring learners to justify or critique AI suggestions\u0026mdash;there is a risk that automated feedback might supplant rather than support deeper processing.\u003c/p\u003e \u003cp\u003eImportantly, these challenges did not diminish AI\u0026rsquo;s overall value but rather pointed to the necessity of a blended, multimodal ecosystem. Comparative studies have shown that no single tool sufficiently addresses every pedagogical need (Stošić \u0026amp; Guill\u0026eacute;n-G\u0026aacute;mez, 2024; Dokukina \u0026amp; Gumanova, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). In our context, the richest learning emerged when AI‐driven drills were integrated with immersive, human‐mediated activities: live, instructor‐led discussions that contextualized AI practice within broader cultural and communicative frameworks. This combination allowed learners to leverage AI\u0026rsquo;s adaptability while still benefiting from the nuanced corrections and sociocultural insights that only a teacher or native speaker can provide.\u003c/p\u003e \u003cp\u003eEnsuring equitable access further emerged as a critical condition for success. High-speed internet, up‐to‐date hardware, and reliable platforms are prerequisites for seamless AI interactions\u0026mdash;resources that are unevenly distributed across regions and institutions (Azmi \u0026amp; Zakaria, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Wulantari et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Without robust infrastructure investments and clear data‐privacy policies, the promise of AI in language education risks exacerbating existing inequities. Moreover, educators themselves require ongoing professional development to navigate AI tools effectively, interpret their outputs, and integrate them into pedagogical strategies that preserve learner autonomy and critical thinking (Pokrivč\u0026aacute;kov\u0026aacute;, 2019).\u003c/p\u003e \u003cp\u003eTheoretical and practical implications\u003c/p\u003e \u003cdiv id=\"Sec31\" class=\"Section2\"\u003e \u003ch2\u003eFor Researchers:\u003c/h2\u003e \u003cp\u003eThis study advances AI-in-language-education research by using Arabic\u0026mdash;a linguistically and culturally complex case\u0026mdash;to evaluate generative AI tools\u0026rsquo; effectiveness in delivering personalized learning, adaptive feedback, and sustained engagement across diverse learners. It calls for interdisciplinary frameworks uniting pedagogy, AI, and sociolinguistics. Future work should examine long-term outcomes, learner autonomy, and cognitive engagement, while addressing AI\u0026rsquo;s limitations with underrepresented dialects. Scholars can also improve AI models\u0026rsquo; accuracy in translation, cultural nuance detection, and pronunciation support through rigorous evaluation and iterative refinement.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec32\" class=\"Section2\"\u003e \u003ch2\u003eFor Students:\u003c/h2\u003e \u003cp\u003e From a theoretical perspective, our findings highlight generative AI\u0026rsquo;s capacity to scaffold autonomous learning, accelerate vocabulary development, and bolster both oral and written Arabic proficiency\u0026mdash;especially by adapting to complex phonetic and grammatical challenges. Practically, tools like ChatGPT and Duolingo enable learners to reinforce classroom instruction, obtain instant clarification, and gain linguistic confidence. Yet, effective use requires digital literacy: students must critically evaluate AI-generated material and guard against overreliance. Ultimately, AI should serve as a complementary aid\u0026mdash;enhancing human-led instruction and cultural immersion rather than replacing them.\u003c/p\u003e \u003cdiv id=\"Sec33\" class=\"Section3\"\u003e \u003ch2\u003eFor Teachers:\u003c/h2\u003e \u003cp\u003eTheoretically, this study underscores educators\u0026rsquo; pivotal role in mediating AI-driven language learning. Rather than displacing teachers, AI enhances their function as facilitators who interpret system outputs, guide learner reflection, and offer culturally grounded feedback. Practically, it advocates for teacher training that blends digital pedagogy with AI literacy. Educators should learn to embed AI tools in lesson design, tailor instruction using AI-generated diagnostics, and coach students on ethical AI engagement. Blended models\u0026mdash;pairing generative AI with live teaching\u0026mdash;can yield more adaptive, learner-centered classrooms.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec34\" class=\"Section3\"\u003e \u003ch2\u003eFor Decision Makers:\u003c/h2\u003e \u003cp\u003eThis study argues that AI should be integrated into education policy as part of a comprehensive digital ecosystem rather than treated in isolation. Language planners and institutional leaders must adopt evidence-based approaches to embed AI in curricula and assessment inclusively. Practically, they should invest in infrastructure for equitable AI access\u0026mdash;especially in non-native Arabic contexts\u0026mdash;and support AI development that respects linguistic and cultural diversity. Continuous professional development for teachers, alongside clear ethical guidelines and quality standards, is essential to harness AI\u0026rsquo;s benefits while mitigating risks of bias, privacy breaches, and misinformation.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Limitations and Future Research","content":"\u003ch2\u003eLimitations\u003c/h2\u003e\u003cp\u003eWhile offering valuable insights into generative AI\u0026rsquo;s support for non-native Arabic learners, this study has several limitations. First, it relies on self-reported data from only 25 participants, which may restrict generalizability and reflect individual biases, expectations, and digital-literacy levels. Second, its focus on Greek-speaking learners\u0026mdash;though culturally varied\u0026mdash;does not capture the full spectrum of global Arabic students. Third, it addresses only Modern Standard Arabic, overlooking regional dialects prevalent in everyday communication. Finally, it omits teachers\u0026rsquo; and AI developers\u0026rsquo; perspectives, which could have deepened understanding of pedagogical and technological design implications.\u003c/p\u003e\n\u003ch3\u003eFuture Research\u003c/h3\u003e\n\u003cp\u003eFuture research should broaden participant samples to encompass greater linguistic and cultural diversity, enabling comparative analyses of how AI supports Arabic learning across different native-language groups. Studies on AI integration for dialectal Arabic\u0026mdash;addressing regional variants, idioms, and sociolinguistic nuances\u0026mdash;are also needed. Longitudinal designs would illuminate the sustained effects of generative AI on proficiency and motivation. Including teachers\u0026rsquo; and instructional designers\u0026rsquo; perspectives can yield a more comprehensive understanding of AI\u0026rsquo;s curricular integration. Finally, experimental comparisons of AI-supported, blended, and traditional methods will empirically validate the pedagogical value of these technologies in real-world settings.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis article has examined how generative AI transforms Arabic language learning for non-native speakers by shifting pedagogy from rote vocabulary drills to immersive, voice-centric practice. Leveraging platforms such as ChatGPT, Duolingo, and Google Translate, these AI systems create culturally authentic dialogue simulations that reduce learner anxiety and foster sustained engagement. Through real-time, adaptive feedback\u0026mdash;spanning spaced-repetition vocabulary reinforcement, dynamic pronunciation modeling, and context-sensitive grammar scaffolding\u0026mdash;AI tailors instruction to each learner\u0026rsquo;s evolving needs. Performance analytics drive automatic adjustments in task difficulty and pacing, fostering self-regulated study and individualized learning pathways. Yet, persistent limitations\u0026mdash;chiefly dialect misrecognition, occasional semantic inaccuracies, and the risk of overreliance\u0026mdash;underscore that AI cannot fully replicate the nuanced insights of human instruction. The most effective implementations therefore blend AI-driven exercises with instructor-mediated dialogue, ensuring culturally grounded correction, critical reflection, and personalized cultural context. As AI architectures integrate richer dialectal corpora and more sophisticated neural models, they promise to further narrow the gap between vocabulary acquisition and vocal fluency. Ultimately, the strategic fusion of generative AI\u0026rsquo;s scalable adaptability with pedagogical expertise offers a scientifically grounded, culturally responsive framework for achieving meaningful spoken mastery of Modern Standard Arabic and its regional variants.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003cstrong\u003eEthics Approval and Consent to Participate\u003c/strong\u003e \u003cp\u003eThis study was reviewed and approved by the Institutional Review Board (IRB) at XXX (Approval Number: Lang. Feb. 2025/13). All procedures involving human participants were conducted in accordance with institutional guidelines and the ethical standards of the American Psychological Association (APA, 2010) and the Declaration of Helsinki (2013).\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eInformed consent\u003c/strong\u003e \u003cp\u003ewas obtained from all participants prior to participation. Participants were informed of the study\u0026rsquo;s purpose, the voluntary nature of their involvement, and their right to withdraw at any time without penalty. Consent was obtained verbally and documented, as approved by the IRB.\u003c/p\u003e \u003c/p\u003e\u003cp\u003e \u003ch2\u003eHuman Ethics and Consent to Participate Declarations\u003c/h2\u003e \u003cp\u003eAll relevant human ethics and consent procedures were followed. There are no additional declarations beyond those described above.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThis research received no specific grant or financial support from any funding agency, commercial, or not-for-profit sectors.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eA.K., B.H., A.K.E., S.N., and Z.N.K. contributed to the conception and design of the study. A.K., B.H., A.K.E., and S.N. contributed to data collection and preliminary analysis. Z.N.K. supervised the research process and contributed to the interpretation of the findings. A.K. and Z.N.K. wrote the main manuscript text. All authors reviewed, revised, and approved the final manuscript.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eData will be available upon request from the corresponding author\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAl-Bulushi, A. H., \u0026amp; Al‐Issa, A. S. (2017). Playing with the language: Investigating the role of communicative games in an Arab language teaching system. \u003cem\u003eInternational Journal of Instruction\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e, 179\u0026ndash;198. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.12973/iji.2017.10212a\u003c/span\u003e\u003cspan address=\"10.12973/iji.2017.10212a\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlenezi, M. A. K., Mohamed, A. M., \u0026amp; Shaaban, T. S. (2023). Revolutionizing EFL special education: How ChatGPT is transforming the way teachers approach language learning. \u003cem\u003eInternational Journal of Technology and Educational Innovation\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e(2), 5\u0026ndash;23.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlmelhes, S. A. (2023). A review of artificial intelligence adoption in second-language learning. \u003cem\u003eTheory and Practice in Language Studies\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(5), 1259\u0026ndash;1269. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.17507/tpls.1305.21\u003c/span\u003e\u003cspan address=\"10.17507/tpls.1305.21\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlmelhes, S. A. (2024). Gamification for teaching the Arabic language to non-native speakers: A systematic literature review. \u003cem\u003eFrontiers in Education\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e., Article 1371955. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/feduc.2024.1371955\u003c/span\u003e\u003cspan address=\"10.3389/feduc.2024.1371955\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAzmi, N. H., \u0026amp; Zakaria, Z. M. (2020). Gamification among UPSI training teacher during teaching practice in school. \u003cem\u003eInternational Journal of Modern Education\u003c/em\u003e, \u003cem\u003e2\u003c/em\u003e, 56\u0026ndash;67. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.35631/IJMOE.24005\u003c/span\u003e\u003cspan address=\"10.35631/IJMOE.24005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBecker, S. A., Brown, M., Dahlstrom, E., Davis, A., DePaul, K., Diaz, V., \u0026amp; Pomerantz, J. (2018). NMC horizon report: 2018 higher education edition. Educause. Retrieved from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://cit.bnu.edu.cn/docs/2018-09/20180918163624337480.pdf\u003c/span\u003e\u003cspan address=\"https://cit.bnu.edu.cn/docs/2018-09/20180918163624337480.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBjelajac, Ž., Filipović, A., \u0026amp; Stošić, L. (2022). Quis custodiet ipsos custodes: Ethical dilemmas of the KM governed by AI. In M. Stanković \u0026amp; V. Nikolić (Eds.), \u003cem\u003eProceedings of the 4th Virtual International Conference Path to a Knowledge Society: Managing Risks and Innovation\u003c/em\u003e (pp. 353\u0026ndash;360). SASA. Retrieved from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://miteam.mi.sanu.ac.rs/asset/G3sjTM6rPPZThzpnM\u003c/span\u003e\u003cspan address=\"https://miteam.mi.sanu.ac.rs/asset/G3sjTM6rPPZThzpnM\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBjelajac, Ž., Filipović, A. M., \u0026amp; Stošić, L. (2023). Can AI be evil: The criminal capacities of ANI. \u003cem\u003eInternational Journal of Cognitive Research in Science Engineering and Education\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(3), 519\u0026ndash;531. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.23947/2334-8496-2023-11-3-519-531\u003c/span\u003e\u003cspan address=\"10.23947/2334-8496-2023-11-3-519-531\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBurgsteiner, H., Kandlhofer, M., \u0026amp; Steinbauer, G. (2016, March). Irobot: Teaching the basics of artificial intelligence in high schools. \u003cem\u003eProceedings of the AAAI Conference on Artificial Intelligence, 30\u003c/em\u003e(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1609/aaai.v30i1.9864\u003c/span\u003e\u003cspan address=\"10.1609/aaai.v30i1.9864\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eConiam, D. (2014). The linguistic accuracy of chatbots: Usability from an ESL perspective. \u003cem\u003eText \u0026amp; Talk\u003c/em\u003e, \u003cem\u003e34\u003c/em\u003e(5), 545\u0026ndash;567. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1515/text-2014-0018\u003c/span\u003e\u003cspan address=\"10.1515/text-2014-0018\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDokukina, I., \u0026amp; Gumanova, J. (2020). The rise of chatbots: New personal assistants in foreign language learning. \u003cem\u003eProcedia Computer Science\u003c/em\u003e, \u003cem\u003e169\u003c/em\u003e, 542\u0026ndash;546.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFathi, J., Rahimi, M., \u0026amp; Derakhshan, A. (2024). Improving EFL learners\u0026rsquo; speaking skills and willingness to communicate via artificial intelligence-mediated interactions. \u003cem\u003eSystem\u003c/em\u003e, \u003cem\u003e121\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.system.2024.103254\u003c/span\u003e\u003cspan address=\"10.1016/j.system.2024.103254\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Article 103254.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFryer, L. K., \u0026amp; Carpenter, R. (2006). Bots as language learning tools. \u003cem\u003eLanguage Learning \u0026amp; Technology\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e(3), 8\u0026ndash;14. https://dx.doi.org/10.125/44068\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGarc\u0026iacute;a Botero, G., Questier, F., \u0026amp; Zhu, C. (2019). Self-directed language learning in a mobile-assisted, out-of-class context: Do students walk the talk? \u003cem\u003eComputer Assisted Language Learning\u003c/em\u003e, \u003cem\u003e32\u003c/em\u003e(1\u0026ndash;2), 71\u0026ndash;97. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/09588221.2018.1485707\u003c/span\u003e\u003cspan address=\"10.1080/09588221.2018.1485707\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGafni, R., Achituv, D. B., \u0026amp; Rachmani, G. J. (2017). Learning foreign languages using mobile applications. \u003cem\u003eJournal of Information Technology Education: Research\u003c/em\u003e, \u003cem\u003e16\u003c/em\u003e, 301\u0026ndash;317.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrishechko, E. G. (2021). \u003cem\u003eCommunicative strategy of politeness in modern English: Multi-level means of implementation\u003c/em\u003e. EHERPQ.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrishechko, E. G. (2023, April 7). Enhancing language learning with podcasts: Evidence from feedback and practice. In \u003cem\u003eProceedings of the 2nd Pan-Russian Research and Practical Conference on Topical Issues of Intercultural Communication and Foreign Language Teaching Methodology\u003c/em\u003e (pp. 69\u0026ndash;72). AkademLit.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrishechko, E. G. (2024). Analysis of the potential of the Fusion Brain neural network as a tool for visualizing training materials. In \u003cem\u003eProceedings of the 11th International Research and Practical Conference on Topical Issues of Linguistics and Teaching Methods in Business and Professional Communication\u003c/em\u003e (pp. 345\u0026ndash;347). RUDN University.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuill\u0026eacute;n-G\u0026aacute;mez, F. D., Lugones, A., \u0026amp; Mayorga-Fern\u0026aacute;ndez, M. J. (2019). ICT use by pre-service foreign languages teachers according to gender, age and motivation. \u003cem\u003eCogent Education\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003eArticle 1574693. https://doi.org/10.1080/2331186X.2019.1574693\u003c/span\u003e\u003cspan address=\"Article 1574693. 10.1080/2331186X.2019.1574693\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHong, W. C. H. (2023). The impact of ChatGPT on foreign language teaching and learning: Opportunities in education and research. \u003cem\u003eJournal of Educational Technology and Innovation, 5\u003c/em\u003e(1). Retrieved from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://jeti.thewsu.org/index.php/cieti/article/view/103/64\u003c/span\u003e\u003cspan address=\"https://jeti.thewsu.org/index.php/cieti/article/view/103/64\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang, W., Hew, K. F., \u0026amp; Fryer, L. K. (2022). Chatbots for language learning: Are they really useful? A systematic review of chatbot-supported language learning. \u003cem\u003eJournal of Computer Assisted Learning\u003c/em\u003e, \u003cem\u003e38\u003c/em\u003e(1), 237\u0026ndash;257. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/jcal.12610\u003c/span\u003e\u003cspan address=\"10.1111/jcal.12610\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKasneci, E., Se\u0026szlig;ler, K., K\u0026uuml;chemann, S., Bannert, M., Dementieva, D., Fischer, F., \u0026amp; Kasneci, G. (2023). ChatGPT for good? On opportunities and challenges of large language models for education. \u003cem\u003eLearning and Individual Differences\u003c/em\u003e, \u003cem\u003e103\u003c/em\u003e, 102274. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.lindif.2023.102274\u003c/span\u003e\u003cspan address=\"10.1016/j.lindif.2023.102274\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlimova, B. (2020). Benefits of the use of mobile applications for learning a foreign language by elderly population. \u003cem\u003eProcedia Computer Science\u003c/em\u003e, \u003cem\u003e176\u003c/em\u003e, 2184\u0026ndash;2191. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.procs.2020.09.255\u003c/span\u003e\u003cspan address=\"10.1016/j.procs.2020.09.255\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlimova, B., \u0026amp; Pikhart, M. (2019). Cognitive and applied linguistics aspects of using social media: The impact of the use of Facebook on developing writing skills in learning English as a foreign language. \u003cem\u003eEuropean Journal of Investigation in Health Psychology and Education\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e(1), 110\u0026ndash;118. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/ejihpe10010010\u003c/span\u003e\u003cspan address=\"10.3390/ejihpe10010010\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKong, S. C., Cheung, W. M. Y., \u0026amp; Zhang, G. (2021). Evaluation of an artificial intelligence literacy course for university students with diverse study backgrounds. \u003cem\u003eComputers and Education: Artificial Intelligence\u003c/em\u003e, \u003cem\u003e2\u003c/em\u003e, 100026. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.caeai.2021.100026\u003c/span\u003e\u003cspan address=\"10.1016/j.caeai.2021.100026\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKenali, H. M. S., Yusoff, N. M. R. N., Saad, M., Abdullah, N. S. B., H., \u0026amp; Kenali, A. M. S. (2019). The effects of language games on smartphones in developing Arabic speaking skills among non-native speakers. \u003cem\u003eCreative Education\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e, 972\u0026ndash;979. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.4236/ce.2019.105073\u003c/span\u003e\u003cspan address=\"10.4236/ce.2019.105073\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKotob, M. M., \u0026amp; Ibrahim, A. (2019). Gamification: The effect on students\u0026rsquo; motivation and achievement in language learning. \u003cem\u003eJournal of Applied Linguistics \u0026amp; Language Research\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e, 177\u0026ndash;198. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.4995/eurocall.2020.12974\u003c/span\u003e\u003cspan address=\"10.4995/eurocall.2020.12974\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLazar, S., \u0026amp; Guill\u0026eacute;n-G\u0026aacute;mez, F. D. (2024). The potential of IT tools in foreign language acquisition: A comparative assessment. \u003cem\u003eTraining Language and Culture\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e(4), 95\u0026ndash;108. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.22363/2521-442X-2024-8-4-95-108\u003c/span\u003e\u003cspan address=\"10.22363/2521-442X-2024-8-4-95-108\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLoewen, S., Crowther, D., Isbell, D. R., Kim, K. M., Maloney, J., Miller, Z. F., \u0026amp; Rawal, H. (2019). Mobile-assisted language learning: A Duolingo case study. \u003cem\u003eReCALL\u003c/em\u003e, \u003cem\u003e31\u003c/em\u003e(3), 293\u0026ndash;311. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1017/S0958344019000340\u003c/span\u003e\u003cspan address=\"10.1017/S0958344019000340\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLord, G. (2015). I don't know how to use words in Spanish\u0026rsquo;: Rosetta Stone and learner proficiency outcomes. \u003cem\u003eThe Modern Language Journal\u003c/em\u003e, \u003cem\u003e99\u003c/em\u003e(2), 401\u0026ndash;405. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/modl.12234_3\u003c/span\u003e\u003cspan address=\"10.1111/modl.12234_3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLu, X. (2019). An empirical study on the artificial intelligence writing evaluation system in China CET. \u003cem\u003eBig Data\u003c/em\u003e, \u003cem\u003e7\u003c/em\u003e(2), 121\u0026ndash;129. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1089/big.2018.0151\u003c/span\u003e\u003cspan address=\"10.1089/big.2018.0151\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMalyuga, E. N. (2007). \u003cem\u003eFunctional-pragmatic features of speech behavior of British and American participants in intercultural business communication\u003c/em\u003e. RUDN University.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMalyuga, E. N. (2008). A newspaper text: Genre and style paradigm in English interrogative sentences. \u003cem\u003eBulletin of the University of the Russian Academy of Education\u003c/em\u003e, \u003cem\u003e2\u003c/em\u003e, 68\u0026ndash;71.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMalyuga, E. N. (2010). Professionally oriented foreign language teaching at the present stage. \u003cem\u003eIssues of Applied Linguistics\u003c/em\u003e, \u003cem\u003e3\u003c/em\u003e, 83\u0026ndash;97.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMalyuga, E. N. (2019). Emergent trends in English scientific discourse: Issues of research relevance and linguistic identity. \u003cem\u003eTomsk State University Journal of Philology\u003c/em\u003e, \u003cem\u003e58\u003c/em\u003e, 52\u0026ndash;70. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.17223/19986645/58/4\u003c/span\u003e\u003cspan address=\"10.17223/19986645/58/4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMellati, M., Khademi, M., \u0026amp; Abolhassani, M. (2018). Creative interaction in social networks: Multi-synchronous language learning environments. \u003cem\u003eEducation and Information Technologies\u003c/em\u003e, \u003cem\u003e23\u003c/em\u003e, 2053\u0026ndash;2071. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10639-018-9703-9\u003c/span\u003e\u003cspan address=\"10.1007/s10639-018-9703-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMohamed, A. M., Shaaban, T. S., Bakry, S. H., Guill\u0026eacute;n-G\u0026aacute;mez, F. D., \u0026amp; Strzelecki, A. (2024). Empowering the faculty of education students: Applying AI\u0026rsquo;s potential for motivating and enhancing learning. \u003cem\u003eInnovative Higher Education\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s10755-024-09747-z\u003c/span\u003e\u003cspan address=\"10.1007/s10755-024-09747-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNasaruddin, N. (2024). Using ChatGPT in teaching Arabic as a foreign language. \u003cem\u003eArabiyatuna: Jurnal Bahasa Arab\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e(1), 1\u0026ndash;24. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://dx.doi.org/10.29240/jba.v8i1.9413\u003c/span\u003e\u003cspan address=\"10.29240/jba.v8i1.9413\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNg, D. T. K., Leung, J. K. L., Chu, S. K. W., \u0026amp; Qiao, M. S. (2021). Conceptualizing AI literacy: An exploratory review. \u003cem\u003eComputers and Education: Artificial Intelligence\u003c/em\u003e, \u003cem\u003e2\u003c/em\u003e, 100041. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.caeai.2021.100041\u003c/span\u003e\u003cspan address=\"10.1016/j.caeai.2021.100041\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNguyen, T. X. (2024). English majors\u0026rsquo; perceptions of AI tool application in English language learning at tertiary level in Vietnam. \u003cem\u003eJournal of Knowledge Learning and Science Technology\u003c/em\u003e, \u003cem\u003e3\u003c/em\u003e(1), 179\u0026ndash;193. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.60087/jklst.vol3.n1.p193\u003c/span\u003e\u003cspan address=\"10.60087/jklst.vol3.n1.p193\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePanagiotidis, P. (2021). Augmented and mixed reality in language learning. \u003cem\u003eEuropean Journal of Education\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e(2), 27\u0026ndash;43. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.26417/501ibq23c\u003c/span\u003e\u003cspan address=\"10.26417/501ibq23c\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePetrosyan, G. O., \u0026amp; Grishechko, E. G. (2020). Integration of linguistic and didactic approaches in the process of foreign language mastering. \u003cem\u003eModern Humanities Success\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e, 125\u0026ndash;131.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePratama, M. P., Sampelolo, R., \u0026amp; Lura, H. (2023). Revolutionizing education: Harnessing the power of artificial intelligence for personalized learning. \u003cem\u003eKlasikal: Journal of Education Language Teaching and Science\u003c/em\u003e, \u003cem\u003e5\u003c/em\u003e(2), 350\u0026ndash;357. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.52208/klasikal.v5i2.877\u003c/span\u003e\u003cspan address=\"10.52208/klasikal.v5i2.877\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRouffet, C., van Beuningen, C., \u0026amp; de Graaff, R. (2023). Constructive alignment in foreign language curricula: An exploration of teaching and assessment practices in Dutch secondary education. \u003cem\u003eThe Language Learning Journal\u003c/em\u003e, \u003cem\u003e51\u003c/em\u003e(3), 344\u0026ndash;358. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/09571736.2022.2025542\u003c/span\u003e\u003cspan address=\"10.1080/09571736.2022.2025542\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eŞah\u0026iacute;n K\u0026ouml;lemen, C. (2024). Artificial intelligence technologies and ethics in educational processes: Solution suggestions and results. \u003cem\u003eInnoeduca: International Journal of Technology and Educational Innovation\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e(2), 201\u0026ndash;216. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.24310/ijtei.102.2024.19806\u003c/span\u003e\u003cspan address=\"10.24310/ijtei.102.2024.19806\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSalma, N. (2020). Collaborative learning: An effective approach to promote language development. \u003cem\u003eInternational Journal of Social Sciences \u0026amp; Educational Studies\u003c/em\u003e, \u003cem\u003e7\u003c/em\u003e(2), 57\u0026ndash;61. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.23918/ijsses.v7i2p57\u003c/span\u003e\u003cspan address=\"10.23918/ijsses.v7i2p57\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSohail, S. S., Farhat, F., Himeur, Y., Nadeem, M., Madsen, D. \u0026Oslash;., Singh, Y., \u0026amp; Mansoor, W. (2023). Decoding ChatGPT: A taxonomy of existing research, current challenges, and possible future directions. \u003cem\u003eJournal of King Saud University \u0026ndash; Computer and Information Sciences\u003c/em\u003e, \u003cem\u003e25\u003c/em\u003e(8). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003eArticle 101675. https://doi.org/10.1016/j.jksuci.2023.101675\u003c/span\u003e\u003cspan address=\"Article 101675. 10.1016/j.jksuci.2023.101675\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStošić, L., \u0026amp; Janković, A. (2023). The impact of artificial intelligence (AI) on education: Balancing advancements and ethical considerations on human rights. \u003cem\u003ePravo \u0026ndash; Teorija i Praksa\u003c/em\u003e, \u003cem\u003e40\u003c/em\u003e(4), 58\u0026ndash;72. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5937/ptp2304058S\u003c/span\u003e\u003cspan address=\"10.5937/ptp2304058S\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStošić, L., \u0026amp; Mikhailova, O. B. (2023). Leadership and innovativeness as the basis for developing teachers\u0026rsquo; digital competences. \u003cem\u003eRUDN Journal of Psychology and Pedagogics\u003c/em\u003e, \u003cem\u003e20\u003c/em\u003e(1), 126\u0026ndash;144. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.22363/2313-1683-2023-20-1-126-144\u003c/span\u003e\u003cspan address=\"10.22363/2313-1683-2023-20-1-126-144\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTalan, T., Doğan, Y., \u0026amp; Kalinkara, Y. (2024). Digital natives\u0026rsquo; mobile learning adoption in terms of UTAUT-2 Model: A structural equation model. \u003cem\u003eInnoeduca: International Journal of Technology and Educational Innovation\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e(1), 100\u0026ndash;123. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.24310/ijtei.101.2024.17440\u003c/span\u003e\u003cspan address=\"10.24310/ijtei.101.2024.17440\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eValencia, E. M., Rivas, E. S., Palmero, J. R., \u0026amp; G\u0026aacute;mez, F. D. G. (2022). Perceptions of university students about virtual reality as a didactic resource: A pre-experimental study with a control and experimental group. \u003cem\u003eIJERI: International Journal of Educational Research and Innovation\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e, 152\u0026ndash;171. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.46661/ijeri.5904\u003c/span\u003e\u003cspan address=\"10.46661/ijeri.5904\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eViktoria, D., Polina, L., Natalia, A., Lilia, N., \u0026amp; Evgenia, E. (2018, September). Virtual and augmented reality in language acquisition. In \u003cem\u003eProceedings of the International Conference on the Theory and Practice of Personality Formation in Modern Society (ICTPPFMS 2018)\u003c/em\u003e (pp. 218\u0026ndash;223). Atlantis Press. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.2991/ictppfms-18.2018.38\u003c/span\u003e\u003cspan address=\"10.2991/ictppfms-18.2018.38\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWoo, J. H., \u0026amp; Choi, H. (2021). Systematic review for AI-based language learning tools. \u003cem\u003eArXiv, 2111.04455\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.48550/arXiv.2111.04455\u003c/span\u003e\u003cspan address=\"10.48550/arXiv.2111.04455\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWulantari, N. P., Rachman, A., Sari, M. N., Uktolseja, L. J., \u0026amp; Rofi\u0026rsquo;i, A. (2023). The role of gamification in English language teaching: A literature review. \u003cem\u003eJournal on Education\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(1), 2847\u0026ndash;2856. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.31004/joe.v6i1.3328\u003c/span\u003e\u003cspan address=\"10.31004/joe.v6i1.3328\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, X., Romero-Forteza, F., \u0026amp; Li, Z. (2023). Bibliometric analysis of technology-supported language learning: LMOOC trend in China. \u003cem\u003eInnoeduca: International Journal of Technology and Educational Innovation\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e(2), 68\u0026ndash;81. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.24310/innoeduca.2023.v9i2.17362\u003c/span\u003e\u003cspan address=\"10.24310/innoeduca.2023.v9i2.17362\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao, X., \u0026amp; Zhang, R. (2019). A state-of-the-art review of automated writing evaluation systems for language learning. \u003cem\u003eComputer Assisted Language Learning\u003c/em\u003e, \u003cem\u003e35\u003c/em\u003e(9), 2790\u0026ndash;2816. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1080/09588221.2021.1896555\u003c/span\u003e\u003cspan address=\"10.1080/09588221.2021.1896555\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Generative AI, Arabic language learning, Non-native speakers, Adaptive feedback, Cultural engagement","lastPublishedDoi":"10.21203/rs.3.rs-9249983/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9249983/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis study examines how generative AI tools reshape Arabic learning for non-native adult speakers, focusing on a diverse cohort of Greek-speaking learners distributed across Saudi Arabia, the UAE, Greece, and other European countries. Through an open-ended questionnaire and thematic analysis, we explored learners\u0026rsquo; experiences, adaptive support features, and perceived challenges associated with AI-driven applications such as ChatGPT, Duolingo, and Google Translate. Findings reveal that AI significantly enhances motivation by simulating culturally rich, low-anxiety conversational scenarios and delivers real-time personalization in vocabulary reinforcement, pronunciation practice, and grammar scaffolding. However, persistent limitations include translation inaccuracies, inadequate handling of regional dialects, and risk of cognitive overreliance when learners depend on AI for routine corrections. The most effective learning environments combined generative AI with instructor-mediated discussions, leveraging AI\u0026rsquo;s strengths while preserving human oversight for nuanced cultural and linguistic guidance. Implications underscore the need for dialect-aware AI models, equitable infrastructure investment, and continuous educator training. This research offers empirical insights for ethically integrating AI into hybrid and virtual Arabic programs to support personalized and culturally responsive language acquisition.\u003c/p\u003e","manuscriptTitle":"From Vocabulary to Voice: How Generative AI Shapes Arabic Learning for Non-Native Speakers","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-07 16:42:22","doi":"10.21203/rs.3.rs-9249983/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"4dfff95b-5da6-46fb-adb4-8f61208675f9","owner":[],"postedDate":"April 7th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Rejected","date":"2026-05-14T01:54:33+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-03T02:29:25+00:00","index":18,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-05-14T02:09:31+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-07 16:42:22","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9249983","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9249983","identity":"rs-9249983","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}